<?xml version="1.0" encoding="UTF-8"?>
<?xml-stylesheet href="/stylesheet.xsl" type="text/xsl"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:sy="http://purl.org/rss/1.0/modules/syndication/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" xmlns:podcast="https://podcastindex.org/namespace/1.0">
  <channel>
    <atom:link rel="self" type="application/rss+xml" href="https://feeds.transistor.fm/the-power-allocation" title="MP3 Audio"/>
    <atom:link rel="hub" href="https://pubsubhubbub.appspot.com/"/>
    <podcast:podping usesPodping="true"/>
    <title>The Power Allocation</title>
    <generator>Transistor (https://transistor.fm)</generator>
    <itunes:new-feed-url>https://feeds.transistor.fm/the-power-allocation</itunes:new-feed-url>
    <description>The AI boom isn't constrained by chips, algorithms, or talent. It's constrained by electricity.

The Power Allocation is a daily briefing on AI infrastructure — where capital is actually being deployed. Each episode cuts through the hype to examine the physical realities shaping the AI buildout: power constraints, grid interconnection, land acquisition, data center financing, cooling infrastructure, and utility relationships.

This isn't a software podcast. This is an infrastructure podcast.

Who it's for: Institutional investors, infrastructure allocators, data center developers, utilities, family offices, and anyone positioning capital for the physical layer of artificial intelligence.

What you'll learn:

Why power availability — not GPU supply — is the binding constraint on AI compute
How hyperscalers are locking in multi-decade power purchase agreements
Where data centers are relocating and why grid geography is reshaping the industry
The financing structures turning compute facilities into bond-like assets
What execution timelines, permitting delays, and interconnection queues mean for capital deployment
Format: 3-6 minute episodes. Dense. Clear. No hype.

New episodes daily. Subscribe wherever you listen.</description>
    <copyright>© 2026 Spring Street Management Group</copyright>
    <podcast:guid>d315076b-b596-50b1-925e-e90a7e5ab61f</podcast:guid>
    <podcast:locked>yes</podcast:locked>
    <language>en</language>
    <pubDate>Fri, 20 Mar 2026 06:00:17 -0700</pubDate>
    <lastBuildDate>Fri, 20 Mar 2026 06:01:57 -0700</lastBuildDate>
    <image>
      <url>https://img.transistorcdn.com/DetXsTyUANDpVGXFWAqrJhIEwWQEFKQgHJVDL5vn2DM/rs:fill:0:0:1/w:1400/h:1400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS80NTY2/MmFkNGZjMzYyZTE1/YWMyMDdiMjdmZjcx/MTcwMi5wbmc.jpg</url>
      <title>The Power Allocation</title>
    </image>
    <itunes:category text="Business"/>
    <itunes:category text="Business">
      <itunes:category text="Investing"/>
    </itunes:category>
    <itunes:type>episodic</itunes:type>
    <itunes:author>Spring Street Management Group</itunes:author>
    <itunes:image href="https://img.transistorcdn.com/DetXsTyUANDpVGXFWAqrJhIEwWQEFKQgHJVDL5vn2DM/rs:fill:0:0:1/w:1400/h:1400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS80NTY2/MmFkNGZjMzYyZTE1/YWMyMDdiMjdmZjcx/MTcwMi5wbmc.jpg"/>
    <itunes:summary>The AI boom isn't constrained by chips, algorithms, or talent. It's constrained by electricity.

The Power Allocation is a daily briefing on AI infrastructure — where capital is actually being deployed. Each episode cuts through the hype to examine the physical realities shaping the AI buildout: power constraints, grid interconnection, land acquisition, data center financing, cooling infrastructure, and utility relationships.

This isn't a software podcast. This is an infrastructure podcast.

Who it's for: Institutional investors, infrastructure allocators, data center developers, utilities, family offices, and anyone positioning capital for the physical layer of artificial intelligence.

What you'll learn:

Why power availability — not GPU supply — is the binding constraint on AI compute
How hyperscalers are locking in multi-decade power purchase agreements
Where data centers are relocating and why grid geography is reshaping the industry
The financing structures turning compute facilities into bond-like assets
What execution timelines, permitting delays, and interconnection queues mean for capital deployment
Format: 3-6 minute episodes. Dense. Clear. No hype.

New episodes daily. Subscribe wherever you listen.</itunes:summary>
    <itunes:subtitle>The AI boom isn't constrained by chips, algorithms, or talent.</itunes:subtitle>
    <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
    <itunes:owner>
      <itunes:name>Spring Street Management Group</itunes:name>
    </itunes:owner>
    <itunes:complete>No</itunes:complete>
    <itunes:explicit>No</itunes:explicit>
    <item>
      <title>Europe's 176 Billion Euro Bet</title>
      <itunes:episode>50</itunes:episode>
      <podcast:episode>50</podcast:episode>
      <itunes:title>Europe's 176 Billion Euro Bet</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">6418b140-2a63-4163-aacb-d797e8762c3e</guid>
      <link>https://share.transistor.fm/s/41fbf6f8</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we analyze the European Data Centre Association's 2026 forecast: 176 billion euros in investment through 2031—and why capacity growth is constrained by grid readiness, not capital availability.

<p>Europe's data center industry is hitting a wall that money alone can't solve. When grid constraints bite, capital flows to whoever has power access.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Frankfurt, Amsterdam, Dublin, London, and Paris have projects stacked waiting for power</li>
<li>atNorth's 300MW mega site in Sollefteå, Sweden—chasing Nordic grid headroom</li>
<li>Vantage's 400MW Bordeaux deal through direct utility partnership</li>
<li>Edinburgh's rejection of a 213MW proposal on environmental and grid grounds</li>
<li>How demand growing 20-25% annually meets grid growth in single digits</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> European data center, EDCA forecast, Frankfurt data center, Amsterdam data center, Nordic data center, atNorth, Vantage Data Centers, grid constraints Europe, power access, Dublin data center</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we analyze the European Data Centre Association's 2026 forecast: 176 billion euros in investment through 2031—and why capacity growth is constrained by grid readiness, not capital availability.

<p>Europe's data center industry is hitting a wall that money alone can't solve. When grid constraints bite, capital flows to whoever has power access.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Frankfurt, Amsterdam, Dublin, London, and Paris have projects stacked waiting for power</li>
<li>atNorth's 300MW mega site in Sollefteå, Sweden—chasing Nordic grid headroom</li>
<li>Vantage's 400MW Bordeaux deal through direct utility partnership</li>
<li>Edinburgh's rejection of a 213MW proposal on environmental and grid grounds</li>
<li>How demand growing 20-25% annually meets grid growth in single digits</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> European data center, EDCA forecast, Frankfurt data center, Amsterdam data center, Nordic data center, atNorth, Vantage Data Centers, grid constraints Europe, power access, Dublin data center</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Fri, 20 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/41fbf6f8/be4cce18.mp3" length="1538089" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>185</itunes:duration>
      <itunes:summary>Europe plans €176B in data center investment through 2031, but grid constraints are forcing capital to chase power access over traditional hubs.</itunes:summary>
      <itunes:subtitle>Europe plans €176B in data center investment through 2031, but grid constraints are forcing capital to chase power access over traditional hubs.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The X-Energy NRC Setback</title>
      <itunes:episode>53</itunes:episode>
      <podcast:episode>53</podcast:episode>
      <itunes:title>The X-Energy NRC Setback</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">d762e1e8-cbeb-4c02-b423-9e598156a354</guid>
      <link>https://share.transistor.fm/s/6acc6836</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we examine the Nuclear Regulatory Commission's intervention in Dow and X-energy's four-reactor SMR proposal in Texas—and what financial qualification concerns mean for the broader advanced nuclear industry.

<p>This matters for data centers because X-energy has agreements with Amazon and others. Any delays in the Dow project ripple through the entire commercialization timeline.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why the Xe-100 reactor at Dow's Calhoun County site was supposed to be the easy case</li>
<li>What NRC financial qualification requirements mean for startup nuclear companies</li>
<li>How X-energy's Amazon agreements depend on proving commercial viability</li>
<li>The broader lesson: advanced nuclear is a financing challenge, not just technology</li>
<li>What X-energy's next capital raise will signal about the Xe-100's future</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> X-energy, Xe-100 reactor, NRC financial qualification, Dow Chemical nuclear, SMR Texas, Amazon nuclear, advanced nuclear financing, TRISO fuel, data center nuclear setback</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we examine the Nuclear Regulatory Commission's intervention in Dow and X-energy's four-reactor SMR proposal in Texas—and what financial qualification concerns mean for the broader advanced nuclear industry.

<p>This matters for data centers because X-energy has agreements with Amazon and others. Any delays in the Dow project ripple through the entire commercialization timeline.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why the Xe-100 reactor at Dow's Calhoun County site was supposed to be the easy case</li>
<li>What NRC financial qualification requirements mean for startup nuclear companies</li>
<li>How X-energy's Amazon agreements depend on proving commercial viability</li>
<li>The broader lesson: advanced nuclear is a financing challenge, not just technology</li>
<li>What X-energy's next capital raise will signal about the Xe-100's future</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> X-energy, Xe-100 reactor, NRC financial qualification, Dow Chemical nuclear, SMR Texas, Amazon nuclear, advanced nuclear financing, TRISO fuel, data center nuclear setback</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Thu, 19 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/6acc6836/864945e7.mp3" length="1370901" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>164</itunes:duration>
      <itunes:summary>The NRC flagged financial concerns on X-energy's Dow SMR project in Texas—a speed bump with ripple effects for Amazon and data center nuclear.</itunes:summary>
      <itunes:subtitle>The NRC flagged financial concerns on X-energy's Dow SMR project in Texas—a speed bump with ripple effects for Amazon and data center nuclear.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>SMR Costs and the $182 Question</title>
      <itunes:episode>56</itunes:episode>
      <podcast:episode>56</podcast:episode>
      <itunes:title>SMR Costs and the $182 Question</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">f4a67915-f191-489e-b15d-3db3af0ec0fd</guid>
      <link>https://share.transistor.fm/s/07b7d8d7</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we examine the economics of small modular reactors: current estimates put SMR electricity at around $182 per megawatt hour—roughly three times solar, four times natural gas. The economics don't work yet. But data centers may change that equation.

<p>When your alternative is no power at all—or waiting five years for interconnection—$182/MWh starts looking reasonable.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why first-of-a-kind reactor costs don't reflect mature technology economics</li>
<li>The troubled history of nuclear cost projections and the industry's credibility problem</li>
<li>How hyperscalers can tolerate premiums that other customers won't pay</li>
<li>Long-term contracts that average costs: $200/MWh year one, $80 by year fifteen</li>
<li>The coordination problem: everyone waiting for someone else to go first</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> SMR cost, $182 per MWh, nuclear economics, first-of-a-kind reactor, data center power premium, nuclear cost projections, SMR scale economies, hyperscaler nuclear investment</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we examine the economics of small modular reactors: current estimates put SMR electricity at around $182 per megawatt hour—roughly three times solar, four times natural gas. The economics don't work yet. But data centers may change that equation.

<p>When your alternative is no power at all—or waiting five years for interconnection—$182/MWh starts looking reasonable.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why first-of-a-kind reactor costs don't reflect mature technology economics</li>
<li>The troubled history of nuclear cost projections and the industry's credibility problem</li>
<li>How hyperscalers can tolerate premiums that other customers won't pay</li>
<li>Long-term contracts that average costs: $200/MWh year one, $80 by year fifteen</li>
<li>The coordination problem: everyone waiting for someone else to go first</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> SMR cost, $182 per MWh, nuclear economics, first-of-a-kind reactor, data center power premium, nuclear cost projections, SMR scale economies, hyperscaler nuclear investment</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Wed, 18 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/07b7d8d7/861efdfb.mp3" length="1516357" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>182</itunes:duration>
      <itunes:summary>SMR electricity costs ~$182/MWh—3x solar, 4x gas. The economics don't work yet, but data centers may pay premiums others won't.</itunes:summary>
      <itunes:subtitle>SMR electricity costs ~$182/MWh—3x solar, 4x gas. The economics don't work yet, but data centers may pay premiums others won't.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The White House Grid Pledge</title>
      <itunes:episode>49</itunes:episode>
      <podcast:episode>49</podcast:episode>
      <itunes:title>The White House Grid Pledge</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">9f44c199-18ad-4c6d-ae70-89228385e8dc</guid>
      <link>https://share.transistor.fm/s/d123ec92</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we break down the March 4th White House announcement where seven major hyperscalers—Microsoft, Google, Amazon, and four others—signed a pledge to fund U.S. power grid upgrades specifically for data centers.

<p>This isn't a press release. It's a structural shift in who pays for grid infrastructure, moving from ratepayer-funded utility builds to direct private capital investment.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why hyperscalers are writing billion-dollar checks for transmission and substations</li>
<li>How the pledge covers interconnection queue acceleration and clean energy commitments</li>
<li>The unusual alignment of private interest and public good in grid upgrades</li>
<li>Why projects that waited decades for approval now have private capital pushing forward</li>
<li>What this means for European markets facing even more severe grid constraints</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> White House data center, grid upgrade, hyperscaler power, Microsoft grid investment, Google power infrastructure, Amazon data center power, transmission upgrade, interconnection queue, clean energy data center</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we break down the March 4th White House announcement where seven major hyperscalers—Microsoft, Google, Amazon, and four others—signed a pledge to fund U.S. power grid upgrades specifically for data centers.

<p>This isn't a press release. It's a structural shift in who pays for grid infrastructure, moving from ratepayer-funded utility builds to direct private capital investment.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why hyperscalers are writing billion-dollar checks for transmission and substations</li>
<li>How the pledge covers interconnection queue acceleration and clean energy commitments</li>
<li>The unusual alignment of private interest and public good in grid upgrades</li>
<li>Why projects that waited decades for approval now have private capital pushing forward</li>
<li>What this means for European markets facing even more severe grid constraints</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> White House data center, grid upgrade, hyperscaler power, Microsoft grid investment, Google power infrastructure, Amazon data center power, transmission upgrade, interconnection queue, clean energy data center</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Tue, 17 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/d123ec92/2e65a232.mp3" length="1436314" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>172</itunes:duration>
      <itunes:summary>Seven hyperscalers signed a White House pledge to fund U.S. power grid upgrades for data centers—a structural shift in who pays for infrastructure.</itunes:summary>
      <itunes:subtitle>Seven hyperscalers signed a White House pledge to fund U.S. power grid upgrades for data centers—a structural shift in who pays for infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Google's Geothermal Expansion</title>
      <itunes:episode>54</itunes:episode>
      <podcast:episode>54</podcast:episode>
      <itunes:title>Google's Geothermal Expansion</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">10e24d12-83c3-4e79-b059-321468d75b4a</guid>
      <link>https://share.transistor.fm/s/cd12aa01</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we analyze Google's deal with Ormat Technologies and NV Energy for up to 150 megawatts of new geothermal capacity in Nevada—and why geothermal is emerging as a strategic power source for hyperscalers.

<p>Geothermal has a unique advantage: it's baseload. Unlike solar and wind, geothermal runs 24/7 at consistent output—matching data center load profiles almost perfectly.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why the Ormat deal is for new capacity, not existing plants—genuine additionality</li>
<li>Nevada's Basin and Range geology and streamlined geothermal permitting</li>
<li>Google's Fervo Energy partnership: 115MW of enhanced geothermal using oil &amp; gas drilling techniques</li>
<li>How Google is approaching 300MW of geothermal—enough for a medium-sized campus</li>
<li>The geographic limits of geothermal and where it makes strategic sense</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Google geothermal, Ormat Technologies, NV Energy, Fervo Energy, enhanced geothermal, Nevada data center power, baseload renewable, 24/7 clean energy, geothermal data center</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we analyze Google's deal with Ormat Technologies and NV Energy for up to 150 megawatts of new geothermal capacity in Nevada—and why geothermal is emerging as a strategic power source for hyperscalers.

<p>Geothermal has a unique advantage: it's baseload. Unlike solar and wind, geothermal runs 24/7 at consistent output—matching data center load profiles almost perfectly.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why the Ormat deal is for new capacity, not existing plants—genuine additionality</li>
<li>Nevada's Basin and Range geology and streamlined geothermal permitting</li>
<li>Google's Fervo Energy partnership: 115MW of enhanced geothermal using oil &amp; gas drilling techniques</li>
<li>How Google is approaching 300MW of geothermal—enough for a medium-sized campus</li>
<li>The geographic limits of geothermal and where it makes strategic sense</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Google geothermal, Ormat Technologies, NV Energy, Fervo Energy, enhanced geothermal, Nevada data center power, baseload renewable, 24/7 clean energy, geothermal data center</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Mon, 16 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/cd12aa01/cb2c6774.mp3" length="1462439" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>176</itunes:duration>
      <itunes:summary>Google signed for 150MW of new geothermal in Nevada with Ormat—adding to Fervo and approaching 300MW of 24/7 underground heat for data centers.</itunes:summary>
      <itunes:subtitle>Google signed for 150MW of new geothermal in Nevada with Ormat—adding to Fervo and approaching 300MW of 24/7 underground heat for data centers.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Duke Energy's 4.5 Gigawatt Data Center Portfolio</title>
      <itunes:episode>51</itunes:episode>
      <podcast:episode>51</podcast:episode>
      <itunes:title>Duke Energy's 4.5 Gigawatt Data Center Portfolio</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">6de23f34-be5d-4bd8-b21e-bd7dde95ddb2</guid>
      <link>https://share.transistor.fm/s/4b3bd2ad</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we examine Duke Energy's data center contracts reaching 4.5 gigawatts after signing a deal with Microsoft for large-scale AI complexes in North Carolina—an unprecedented concentration of demand from a single customer category.

<p>4.5 gigawatts equals roughly four nuclear power plants. Duke now faces adding generation capacity at a pace not attempted in decades.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Duke's Carolinas territory became ground zero for hyperscale expansion</li>
<li>The challenge of matching 24/7 data center load with intermittent renewables</li>
<li>Duke's "build everything" response: solar, batteries, natural gas, and nuclear exploration</li>
<li>How utilities built for 1-2% annual growth face 10-20% growth in specific service areas</li>
<li>Why solving this puzzle defines American electricity infrastructure for a generation</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Duke Energy, Microsoft data center, North Carolina data center, utility data center, gigawatt portfolio, power generation, clean energy procurement, grid capacity, Charlotte data center, Raleigh data center</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we examine Duke Energy's data center contracts reaching 4.5 gigawatts after signing a deal with Microsoft for large-scale AI complexes in North Carolina—an unprecedented concentration of demand from a single customer category.

<p>4.5 gigawatts equals roughly four nuclear power plants. Duke now faces adding generation capacity at a pace not attempted in decades.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Duke's Carolinas territory became ground zero for hyperscale expansion</li>
<li>The challenge of matching 24/7 data center load with intermittent renewables</li>
<li>Duke's "build everything" response: solar, batteries, natural gas, and nuclear exploration</li>
<li>How utilities built for 1-2% annual growth face 10-20% growth in specific service areas</li>
<li>Why solving this puzzle defines American electricity infrastructure for a generation</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Duke Energy, Microsoft data center, North Carolina data center, utility data center, gigawatt portfolio, power generation, clean energy procurement, grid capacity, Charlotte data center, Raleigh data center</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Fri, 13 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/4b3bd2ad/184f98d5.mp3" length="1623163" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>196</itunes:duration>
      <itunes:summary>Duke Energy hits 4.5GW in data center contracts after Microsoft deal—equivalent to four nuclear plants for one customer category.</itunes:summary>
      <itunes:subtitle>Duke Energy hits 4.5GW in data center contracts after Microsoft deal—equivalent to four nuclear plants for one customer category.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The 44 Gigawatt Gap</title>
      <itunes:episode>55</itunes:episode>
      <podcast:episode>55</podcast:episode>
      <itunes:title>The 44 Gigawatt Gap</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">8cf62884-6168-4a9b-97ce-f102e3bca9ae</guid>
      <link>https://share.transistor.fm/s/8c16b292</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we examine the projected 44 gigawatt shortfall in U.S. electricity supply by 2028—driven primarily by AI data center growth—and why closing it requires the largest generation buildout since rural electrification.

<p>Forty-four gigawatts equals roughly 44 nuclear reactors, or 150 natural gas plants, or an incomprehensible amount of solar and wind with storage.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>The timeline mismatch: data centers need power in 18-24 months; power plants take 5-10 years</li>
<li>Why hyperscalers are pursuing every option simultaneously: nuclear restarts, SMRs, geothermal, behind-the-meter</li>
<li>How Jevons Paradox is overwhelming efficiency gains in AI compute</li>
<li>Why the 44GW figure may be conservative if AI growth accelerates</li>
<li>How electricity is becoming a strategic resource in the AI race</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> 44 gigawatt gap, AI power demand, electricity shortfall, data center energy crisis, power generation buildout, Jevons Paradox AI, hyperscaler power strategy, grid capacity shortage</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we examine the projected 44 gigawatt shortfall in U.S. electricity supply by 2028—driven primarily by AI data center growth—and why closing it requires the largest generation buildout since rural electrification.

<p>Forty-four gigawatts equals roughly 44 nuclear reactors, or 150 natural gas plants, or an incomprehensible amount of solar and wind with storage.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>The timeline mismatch: data centers need power in 18-24 months; power plants take 5-10 years</li>
<li>Why hyperscalers are pursuing every option simultaneously: nuclear restarts, SMRs, geothermal, behind-the-meter</li>
<li>How Jevons Paradox is overwhelming efficiency gains in AI compute</li>
<li>Why the 44GW figure may be conservative if AI growth accelerates</li>
<li>How electricity is becoming a strategic resource in the AI race</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> 44 gigawatt gap, AI power demand, electricity shortfall, data center energy crisis, power generation buildout, Jevons Paradox AI, hyperscaler power strategy, grid capacity shortage</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Fri, 13 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/8c16b292/d7944092.mp3" length="1499627" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>180</itunes:duration>
      <itunes:summary>Multiple analyses project a 44GW U.S. electricity shortfall by 2028 from AI data centers—the largest generation buildout since rural electrification.</itunes:summary>
      <itunes:subtitle>Multiple analyses project a 44GW U.S. electricity shortfall by 2028 from AI data centers—the largest generation buildout since rural electrification.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The White House Grid Pledge</title>
      <itunes:episode>49</itunes:episode>
      <podcast:episode>49</podcast:episode>
      <itunes:title>The White House Grid Pledge</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">e9609519-6212-41fc-b505-4b1978e0b596</guid>
      <link>https://share.transistor.fm/s/2b39c7a4</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we break down the March 4th White House announcement where seven major hyperscalers—Microsoft, Google, Amazon, and others—signed a pledge to fund U.S. power grid upgrades specifically for data centers.

<p>This isn't a press release. It's a structural shift in who pays for grid infrastructure. For decades, utilities built transmission with ratepayer money. Now tech companies are writing checks directly.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why traditional utility planning cycles can't accommodate hyperscaler timelines</li>
<li>What the pledge covers: transmission upgrades, substation expansions, interconnection acceleration</li>
<li>How private grid funding creates unusual alignment of corporate and public interest</li>
<li>The catch: investments flow only where hyperscalers want to build</li>
<li>Why Europe may follow the American experiment in private grid funding</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> White House data center, grid infrastructure, hyperscaler power, Microsoft data center, Google data center, Amazon data center, grid upgrade, transmission infrastructure, power grid investment, utility planning</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we break down the March 4th White House announcement where seven major hyperscalers—Microsoft, Google, Amazon, and others—signed a pledge to fund U.S. power grid upgrades specifically for data centers.

<p>This isn't a press release. It's a structural shift in who pays for grid infrastructure. For decades, utilities built transmission with ratepayer money. Now tech companies are writing checks directly.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why traditional utility planning cycles can't accommodate hyperscaler timelines</li>
<li>What the pledge covers: transmission upgrades, substation expansions, interconnection acceleration</li>
<li>How private grid funding creates unusual alignment of corporate and public interest</li>
<li>The catch: investments flow only where hyperscalers want to build</li>
<li>Why Europe may follow the American experiment in private grid funding</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> White House data center, grid infrastructure, hyperscaler power, Microsoft data center, Google data center, Amazon data center, grid upgrade, transmission infrastructure, power grid investment, utility planning</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Thu, 12 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/2b39c7a4/604de8ae.mp3" length="1436314" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>172</itunes:duration>
      <itunes:summary>Seven hyperscalers sign White House pledge to fund U.S. power grid upgrades—a structural shift in who pays for infrastructure.</itunes:summary>
      <itunes:subtitle>Seven hyperscalers sign White House pledge to fund U.S. power grid upgrades—a structural shift in who pays for infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>SMR Costs and the $182 Question</title>
      <itunes:episode>56</itunes:episode>
      <podcast:episode>56</podcast:episode>
      <itunes:title>SMR Costs and the $182 Question</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">5a17fcf2-a3db-4c3b-bf56-9ccdc55579c4</guid>
      <link>https://share.transistor.fm/s/e42e87e9</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we examine the economic reality of small modular reactors for data centers: current estimates put SMR electricity at around $182 per megawatt hour—roughly three times utility-scale solar and four times combined cycle natural gas.

<p>The SMR value proposition depends on costs falling dramatically through factory fabrication and standardized designs. But data centers may tolerate premium pricing when the alternative is no power at all.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why first-of-a-kind reactors carry full design, regulatory, and supply chain development costs</li>
<li>The nuclear industry's credibility problem after Vogtle cost overruns</li>
<li>How hyperscalers' long-term contracts can average costs over decades</li>
<li>The coordination problem: everyone waiting for someone else to prove costs come down</li>
<li>Why $182/MWh looks reasonable when the alternative is 5-year grid interconnection waits</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> SMR costs, small modular reactor economics, nuclear power costs, LCOE, data center nuclear, Vogtle, factory fabrication, nuclear economics, energy costs, power purchase agreement, hyperscaler nuclear</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we examine the economic reality of small modular reactors for data centers: current estimates put SMR electricity at around $182 per megawatt hour—roughly three times utility-scale solar and four times combined cycle natural gas.

<p>The SMR value proposition depends on costs falling dramatically through factory fabrication and standardized designs. But data centers may tolerate premium pricing when the alternative is no power at all.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why first-of-a-kind reactors carry full design, regulatory, and supply chain development costs</li>
<li>The nuclear industry's credibility problem after Vogtle cost overruns</li>
<li>How hyperscalers' long-term contracts can average costs over decades</li>
<li>The coordination problem: everyone waiting for someone else to prove costs come down</li>
<li>Why $182/MWh looks reasonable when the alternative is 5-year grid interconnection waits</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> SMR costs, small modular reactor economics, nuclear power costs, LCOE, data center nuclear, Vogtle, factory fabrication, nuclear economics, energy costs, power purchase agreement, hyperscaler nuclear</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Thu, 12 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/e42e87e9/4473333f.mp3" length="1516357" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>182</itunes:duration>
      <itunes:summary>U.S. bets $9B on SMRs for data centers, but $182/MWh cost estimates are 3x solar and 4x natural gas. Can scale economics change the math?</itunes:summary>
      <itunes:subtitle>U.S. bets $9B on SMRs for data centers, but $182/MWh cost estimates are 3x solar and 4x natural gas. Can scale economics change the math?</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Europe's 176 Billion Euro Bet</title>
      <itunes:episode>50</itunes:episode>
      <podcast:episode>50</podcast:episode>
      <itunes:title>Europe's 176 Billion Euro Bet</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">8901bc3b-7b30-443a-be63-f0f9d0d49133</guid>
      <link>https://share.transistor.fm/s/6b244f3d</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we analyze the European Data Centre Association's 2026 forecast projecting 176 billion euros in investment through 2031—and the grid constraints that determine where that capital can actually flow.

<p>Europe's data center industry is hitting a wall that money alone can't solve. Frankfurt, Amsterdam, Dublin, London, Paris—every major hub has projects waiting for power connections that utilities can't deliver.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why demand growing 20-25% annually versus single-digit grid capacity growth creates compounding gaps</li>
<li>atNorth's 300MW Sweden mega site and Vantage's 400MW Bordeaux deal as adaptation strategies</li>
<li>Edinburgh's 213MW rejection signals growing community and grid resistance</li>
<li>How power access is reshaping European data center geography beyond traditional hubs</li>
<li>Why developers now lead with power analysis, not real estate analysis</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> European data center, EDCA, Frankfurt data center, Amsterdam data center, Dublin data center, grid constraints, atNorth, Vantage Data Centers, Nordic data center, power infrastructure Europe</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we analyze the European Data Centre Association's 2026 forecast projecting 176 billion euros in investment through 2031—and the grid constraints that determine where that capital can actually flow.

<p>Europe's data center industry is hitting a wall that money alone can't solve. Frankfurt, Amsterdam, Dublin, London, Paris—every major hub has projects waiting for power connections that utilities can't deliver.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why demand growing 20-25% annually versus single-digit grid capacity growth creates compounding gaps</li>
<li>atNorth's 300MW Sweden mega site and Vantage's 400MW Bordeaux deal as adaptation strategies</li>
<li>Edinburgh's 213MW rejection signals growing community and grid resistance</li>
<li>How power access is reshaping European data center geography beyond traditional hubs</li>
<li>Why developers now lead with power analysis, not real estate analysis</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> European data center, EDCA, Frankfurt data center, Amsterdam data center, Dublin data center, grid constraints, atNorth, Vantage Data Centers, Nordic data center, power infrastructure Europe</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Wed, 11 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/6b244f3d/53e40697.mp3" length="1538089" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>185</itunes:duration>
      <itunes:summary>European Data Centre Association forecasts €176B investment through 2031—but grid readiness constrains where capital can actually deploy.</itunes:summary>
      <itunes:subtitle>European Data Centre Association forecasts €176B investment through 2031—but grid readiness constrains where capital can actually deploy.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Duke Energy's 4.5 Gigawatt Data Center Portfolio</title>
      <itunes:episode>51</itunes:episode>
      <podcast:episode>51</podcast:episode>
      <itunes:title>Duke Energy's 4.5 Gigawatt Data Center Portfolio</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">5740d4e5-3954-4f24-8751-c993a43cc419</guid>
      <link>https://share.transistor.fm/s/a798940a</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we examine Duke Energy's unprecedented data center portfolio: 4.5 gigawatts of contracted capacity after signing a deal with Microsoft for large-scale AI complexes in North Carolina.

<p>To put that in perspective: 4.5 gigawatts equals roughly four nuclear power plants. One utility. One customer category. That concentration of demand is unprecedented in American electricity history.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Duke's Carolinas, Indiana, and Florida service territory became hyperscale ground zero</li>
<li>The generation challenge: extending coal retirements, building gas, exploring nuclear—simultaneously</li>
<li>How clean energy procurement requirements clash with 24/7 data center load profiles</li>
<li>The math problem of matching intermittent renewables to constant demand</li>
<li>Why utilities built for 1-2% growth now face 10-20% growth in specific areas</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Duke Energy data center, Microsoft North Carolina, 4.5 gigawatt, utility data center contract, clean energy procurement, hyperscale power, Charlotte data center, Raleigh data center, grid capacity</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we examine Duke Energy's unprecedented data center portfolio: 4.5 gigawatts of contracted capacity after signing a deal with Microsoft for large-scale AI complexes in North Carolina.

<p>To put that in perspective: 4.5 gigawatts equals roughly four nuclear power plants. One utility. One customer category. That concentration of demand is unprecedented in American electricity history.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Duke's Carolinas, Indiana, and Florida service territory became hyperscale ground zero</li>
<li>The generation challenge: extending coal retirements, building gas, exploring nuclear—simultaneously</li>
<li>How clean energy procurement requirements clash with 24/7 data center load profiles</li>
<li>The math problem of matching intermittent renewables to constant demand</li>
<li>Why utilities built for 1-2% growth now face 10-20% growth in specific areas</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Duke Energy data center, Microsoft North Carolina, 4.5 gigawatt, utility data center contract, clean energy procurement, hyperscale power, Charlotte data center, Raleigh data center, grid capacity</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Wed, 11 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/a798940a/840f840e.mp3" length="1623163" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>196</itunes:duration>
      <itunes:summary>Duke Energy's data center contracts hit 4.5GW after Microsoft's North Carolina deal—the equivalent of four nuclear plants for one customer category.</itunes:summary>
      <itunes:subtitle>Duke Energy's data center contracts hit 4.5GW after Microsoft's North Carolina deal—the equivalent of four nuclear plants for one customer category.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Meta's Billion-Dollar Indiana Campus</title>
      <itunes:episode>52</itunes:episode>
      <podcast:episode>52</podcast:episode>
      <itunes:title>Meta's Billion-Dollar Indiana Campus</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">b607efb1-57b0-42fb-9174-ddd9ae3ae1d7</guid>
      <link>https://share.transistor.fm/s/df232c0a</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we analyze Meta's groundbreaking on a 1 gigawatt data center campus in Lebanon, Indiana—backed by a $10 billion investment that demonstrates how secondary markets can capture hyperscale infrastructure by solving the power equation.

<p>One gigawatt equals enough power for 750,000 homes. Meta is building that capacity for servers in what was recently farmland.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Indiana won: lower costs, available land, cooperative utility (AES Indiana), aggressive incentives</li>
<li>What $10B actually buys: substations, transmission, backup generation, cooling—an industrial city</li>
<li>How Meta's 6.6GW nuclear procurement announcements connect to this campus</li>
<li>The new economic development playbook: power availability and permitting speed over traditional incentives</li>
<li>Why states with constrained grids weren't even in the running</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Meta data center, Indiana data center, Lebanon Indiana, hyperscale data center, AES Indiana, gigawatt campus, data center investment, economic development, Midwest data center, Meta AI infrastructure</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we analyze Meta's groundbreaking on a 1 gigawatt data center campus in Lebanon, Indiana—backed by a $10 billion investment that demonstrates how secondary markets can capture hyperscale infrastructure by solving the power equation.

<p>One gigawatt equals enough power for 750,000 homes. Meta is building that capacity for servers in what was recently farmland.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Indiana won: lower costs, available land, cooperative utility (AES Indiana), aggressive incentives</li>
<li>What $10B actually buys: substations, transmission, backup generation, cooling—an industrial city</li>
<li>How Meta's 6.6GW nuclear procurement announcements connect to this campus</li>
<li>The new economic development playbook: power availability and permitting speed over traditional incentives</li>
<li>Why states with constrained grids weren't even in the running</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Meta data center, Indiana data center, Lebanon Indiana, hyperscale data center, AES Indiana, gigawatt campus, data center investment, economic development, Midwest data center, Meta AI infrastructure</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Tue, 10 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/df232c0a/1e1d6e90.mp3" length="1377391" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>165</itunes:duration>
      <itunes:summary>Meta breaks ground on 1GW data center campus in Lebanon, Indiana with $10B investment—showing how secondary markets win on power access.</itunes:summary>
      <itunes:subtitle>Meta breaks ground on 1GW data center campus in Lebanon, Indiana with $10B investment—showing how secondary markets win on power access.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Kazakhstan's Gigawatt Ambition</title>
      <itunes:episode>57</itunes:episode>
      <podcast:episode>57</podcast:episode>
      <itunes:title>Kazakhstan's Gigawatt Ambition</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">ff0301c8-9892-415d-b88a-800a6fdf6fb8</guid>
      <link>https://share.transistor.fm/s/89db01a7</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we analyze Kazakhstan's announcement of Central Asia's largest data center campus in Ekibastuz with up to 1 gigawatt of energy capacity—and what it signals about AI infrastructure's global migration.

<p>Ekibastuz has what most markets lack: abundant, cheap electricity from Soviet-era coal plants now underutilized. Kazakhstan is positioning itself for customers who prioritize cost over carbon.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>How Soviet-era coal infrastructure becomes a competitive advantage for AI workloads</li>
<li>Target customers: Chinese companies, Russian firms, Indian enterprises, Middle Eastern funds</li>
<li>Why data sovereignty and compute access are becoming strategic national resources</li>
<li>Saudi Arabia, UAE, and Malaysia's parallel moves to capture AI infrastructure</li>
<li>How the global compute map is being redrawn along energy cost lines</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Kazakhstan data center, Ekibastuz, Central Asia data center, global AI infrastructure, data sovereignty, cheap electricity, coal power data center, emerging market data center, compute geography, AI geopolitics</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we analyze Kazakhstan's announcement of Central Asia's largest data center campus in Ekibastuz with up to 1 gigawatt of energy capacity—and what it signals about AI infrastructure's global migration.

<p>Ekibastuz has what most markets lack: abundant, cheap electricity from Soviet-era coal plants now underutilized. Kazakhstan is positioning itself for customers who prioritize cost over carbon.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>How Soviet-era coal infrastructure becomes a competitive advantage for AI workloads</li>
<li>Target customers: Chinese companies, Russian firms, Indian enterprises, Middle Eastern funds</li>
<li>Why data sovereignty and compute access are becoming strategic national resources</li>
<li>Saudi Arabia, UAE, and Malaysia's parallel moves to capture AI infrastructure</li>
<li>How the global compute map is being redrawn along energy cost lines</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Kazakhstan data center, Ekibastuz, Central Asia data center, global AI infrastructure, data sovereignty, cheap electricity, coal power data center, emerging market data center, compute geography, AI geopolitics</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Tue, 10 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/89db01a7/896fc497.mp3" length="1470799" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>177</itunes:duration>
      <itunes:summary>Kazakhstan plans Central Asia's largest data center campus in Ekibastuz with 1GW capacity—a sign of how cheap power reshapes global AI geography.</itunes:summary>
      <itunes:subtitle>Kazakhstan plans Central Asia's largest data center campus in Ekibastuz with 1GW capacity—a sign of how cheap power reshapes global AI geography.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Meta's Billion-Dollar Indiana Campus</title>
      <itunes:episode>52</itunes:episode>
      <podcast:episode>52</podcast:episode>
      <itunes:title>Meta's Billion-Dollar Indiana Campus</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">49af3bd1-281b-4432-82f0-c046c569111e</guid>
      <link>https://share.transistor.fm/s/2dc3005c</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we analyze Meta's groundbreaking on a 1 gigawatt data center campus in Lebanon, Indiana, backed by a $10 billion investment.

<p>One gigawatt. One campus. That's enough power for 750,000 homes—built for servers. Indiana won this project by solving the power equation: available land, cooperative utilities, streamlined permitting, and credible grid capacity.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Lebanon, Indiana beat coastal markets for a $10B hyperscale investment</li>
<li>AES Indiana's role in creative power solutions for massive load</li>
<li>How Meta's 6.6GW nuclear procurement announcements connect to Indiana operations</li>
<li>The infrastructure beyond servers: substations, transmission, cooling at hyperscale</li>
<li>Indiana as a replicable playbook for secondary market data center capture</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Meta data center, Indiana data center, Lebanon Indiana, 1 gigawatt campus, AES Indiana, hyperscale data center, data center investment, economic development, power infrastructure, Meta nuclear</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we analyze Meta's groundbreaking on a 1 gigawatt data center campus in Lebanon, Indiana, backed by a $10 billion investment.

<p>One gigawatt. One campus. That's enough power for 750,000 homes—built for servers. Indiana won this project by solving the power equation: available land, cooperative utilities, streamlined permitting, and credible grid capacity.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Lebanon, Indiana beat coastal markets for a $10B hyperscale investment</li>
<li>AES Indiana's role in creative power solutions for massive load</li>
<li>How Meta's 6.6GW nuclear procurement announcements connect to Indiana operations</li>
<li>The infrastructure beyond servers: substations, transmission, cooling at hyperscale</li>
<li>Indiana as a replicable playbook for secondary market data center capture</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Meta data center, Indiana data center, Lebanon Indiana, 1 gigawatt campus, AES Indiana, hyperscale data center, data center investment, economic development, power infrastructure, Meta nuclear</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Mon, 09 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/2dc3005c/399ed0a9.mp3" length="1377391" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>165</itunes:duration>
      <itunes:summary>Meta breaks ground on 1GW data center campus in Lebanon, Indiana with $10B investment—a case study in how secondary markets capture hyperscale.</itunes:summary>
      <itunes:subtitle>Meta breaks ground on 1GW data center campus in Lebanon, Indiana with $10B investment—a case study in how secondary markets capture hyperscale.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Meta's Billion-Dollar Indiana Campus</title>
      <itunes:episode>52</itunes:episode>
      <podcast:episode>52</podcast:episode>
      <itunes:title>Meta's Billion-Dollar Indiana Campus</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">0c7a444e-6e08-4eea-ad45-18f9a181f2fc</guid>
      <link>https://share.transistor.fm/s/c95e8dd5</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we analyze Meta's groundbreaking on a 1 gigawatt data center campus in Lebanon, Indiana—backed by a $10 billion investment that's reshaping how states compete for hyperscale projects.

<p>One gigawatt. One campus. That's enough to power 750,000 homes—and Meta is building it for servers. Indiana won by solving the power equation.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Lebanon, Indiana beat coastal markets: cheap power, available land, cooperative utility</li>
<li>AES Indiana's creative power solutions and state economic development incentives</li>
<li>Building a small city's worth of electrical infrastructure from scratch</li>
<li>How Meta's 6.6GW nuclear procurement connects to the Indiana campus</li>
<li>The new economic development playbook: power availability over traditional incentives</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Meta data center, Lebanon Indiana, 1 gigawatt campus, $10 billion data center, AES Indiana, hyperscale Midwest, Indiana economic development, Meta nuclear power, secondary market data center</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we analyze Meta's groundbreaking on a 1 gigawatt data center campus in Lebanon, Indiana—backed by a $10 billion investment that's reshaping how states compete for hyperscale projects.

<p>One gigawatt. One campus. That's enough to power 750,000 homes—and Meta is building it for servers. Indiana won by solving the power equation.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Lebanon, Indiana beat coastal markets: cheap power, available land, cooperative utility</li>
<li>AES Indiana's creative power solutions and state economic development incentives</li>
<li>Building a small city's worth of electrical infrastructure from scratch</li>
<li>How Meta's 6.6GW nuclear procurement connects to the Indiana campus</li>
<li>The new economic development playbook: power availability over traditional incentives</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Meta data center, Lebanon Indiana, 1 gigawatt campus, $10 billion data center, AES Indiana, hyperscale Midwest, Indiana economic development, Meta nuclear power, secondary market data center</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Mon, 09 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/c95e8dd5/d8a9e947.mp3" length="1377391" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>165</itunes:duration>
      <itunes:summary>Meta breaks ground on a 1GW, $10B data center campus in Lebanon, Indiana—proof that secondary markets win by solving the power equation.</itunes:summary>
      <itunes:subtitle>Meta breaks ground on a 1GW, $10B data center campus in Lebanon, Indiana—proof that secondary markets win by solving the power equation.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Google's Geothermal Expansion</title>
      <itunes:episode>54</itunes:episode>
      <podcast:episode>54</podcast:episode>
      <itunes:title>Google's Geothermal Expansion</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">05405388-01b0-45f8-9bba-1e76fae8aa7e</guid>
      <link>https://share.transistor.fm/s/c5983082</link>
      <description>
        <![CDATA[In this episode of The Power Allocation, we analyze Google's expanding geothermal strategy following their deal with Ormat Technologies and NV Energy for up to 150 megawatts of new geothermal capacity in Nevada.

<p>Geothermal offers something unique for data centers: true baseload power. Unlike solar and wind, geothermal runs 24/7—matching data center load profiles almost perfectly with no intermittency, no massive battery storage, no backup generation required.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why the Ormat deal for new capacity represents genuine additionality on the grid</li>
<li>How Nevada's Basin and Range geology and streamlined permitting enable expansion</li>
<li>Google's Fervo Energy partnership and enhanced geothermal's broader potential</li>
<li>Why 300MW of geothermal can power a medium-sized data center campus carbon-free</li>
<li>Geographic limitations and why geothermal remains concentrated in specific regions</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Google geothermal, Ormat Technologies, NV Energy, Nevada data center, geothermal power, Fervo Energy, enhanced geothermal, baseload power, clean energy data center, renewable energy, carbon-free power</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[In this episode of The Power Allocation, we analyze Google's expanding geothermal strategy following their deal with Ormat Technologies and NV Energy for up to 150 megawatts of new geothermal capacity in Nevada.

<p>Geothermal offers something unique for data centers: true baseload power. Unlike solar and wind, geothermal runs 24/7—matching data center load profiles almost perfectly with no intermittency, no massive battery storage, no backup generation required.</p>

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why the Ormat deal for new capacity represents genuine additionality on the grid</li>
<li>How Nevada's Basin and Range geology and streamlined permitting enable expansion</li>
<li>Google's Fervo Energy partnership and enhanced geothermal's broader potential</li>
<li>Why 300MW of geothermal can power a medium-sized data center campus carbon-free</li>
<li>Geographic limitations and why geothermal remains concentrated in specific regions</li>
</ul>

<p><strong>About The Power Allocation:</strong> Brought to you by Spring Street Management Group, translating data center and energy hype into real infrastructure and assets on the daily.</p>

<p><strong>Keywords:</strong> Google geothermal, Ormat Technologies, NV Energy, Nevada data center, geothermal power, Fervo Energy, enhanced geothermal, baseload power, clean energy data center, renewable energy, carbon-free power</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Mon, 09 Mar 2026 06:00:00 -0700</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/c5983082/c67754f8.mp3" length="1462439" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>176</itunes:duration>
      <itunes:summary>Google signs deal with Ormat and NV Energy for 150MW new geothermal capacity in Nevada, approaching 300MW total geothermal portfolio.</itunes:summary>
      <itunes:subtitle>Google signs deal with Ormat and NV Energy for 150MW new geothermal capacity in Nevada, approaching 300MW total geothermal portfolio.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Real Timeline for Data Center Nuclear</title>
      <itunes:episode>48</itunes:episode>
      <podcast:episode>48</podcast:episode>
      <itunes:title>The Real Timeline for Data Center Nuclear</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">ea8ed3a4-0f72-44a3-9774-0362904f5dfb</guid>
      <link>https://share.transistor.fm/s/4108029e</link>
      <description>
        <![CDATA[<p>This episode provides a realistic assessment of when nuclear power will actually reach data centers, broken down by pathway.</p><p><strong>Key topics covered:</strong></p><ul><li>Reactor restarts (fastest): Three Mile Island 2028, Palisades 2025</li><li>NuScale certified SMRs: 2028-2030 first deployments</li><li>Advanced designs (Oklo, Kairos, X-energy): 2030+ commercial operation</li><li>Micro reactors: fast construction once NRC certified</li></ul><p><strong>Related keywords:</strong> nuclear timeline data centers, reactor restart timeline, SMR deployment schedule, advanced nuclear 2030, data center power planning</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>This episode provides a realistic assessment of when nuclear power will actually reach data centers, broken down by pathway.</p><p><strong>Key topics covered:</strong></p><ul><li>Reactor restarts (fastest): Three Mile Island 2028, Palisades 2025</li><li>NuScale certified SMRs: 2028-2030 first deployments</li><li>Advanced designs (Oklo, Kairos, X-energy): 2030+ commercial operation</li><li>Micro reactors: fast construction once NRC certified</li></ul><p><strong>Related keywords:</strong> nuclear timeline data centers, reactor restart timeline, SMR deployment schedule, advanced nuclear 2030, data center power planning</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Fri, 06 Mar 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/4108029e/b456e720.mp3" length="1426715" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>171</itunes:duration>
      <itunes:summary>Nuclear will power data centers. The question is when. Here's the realistic assessment of timelines by pathway.</itunes:summary>
      <itunes:subtitle>Nuclear will power data centers. The question is when. Here's the realistic assessment of timelines by pathway.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why Tech Giants Are Buying Nuclear Directly</title>
      <itunes:episode>47</itunes:episode>
      <podcast:episode>47</podcast:episode>
      <itunes:title>Why Tech Giants Are Buying Nuclear Directly</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">0e596a20-765f-46ee-80c6-a480ff53e06a</guid>
      <link>https://share.transistor.fm/s/43e06d0e</link>
      <description>
        <![CDATA[<p>Hyperscalers are no longer waiting for utilities to build nuclear capacity—they're contracting directly with generators and investing billions in new reactor development.</p><p><strong>Key topics covered:</strong></p><ul><li>Microsoft Three Mile Island restart agreement</li><li>Amazon: $20B+ nuclear infrastructure investment including X-energy</li><li>Google-Kairos Power: 500 MW advanced reactor partnership</li><li>Meta: 6.6 GW nuclear agreements</li></ul><p><strong>Related keywords:</strong> hyperscaler nuclear procurement, Microsoft Three Mile Island, Amazon nuclear investment, Google Kairos Power, data center nuclear PPA</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Hyperscalers are no longer waiting for utilities to build nuclear capacity—they're contracting directly with generators and investing billions in new reactor development.</p><p><strong>Key topics covered:</strong></p><ul><li>Microsoft Three Mile Island restart agreement</li><li>Amazon: $20B+ nuclear infrastructure investment including X-energy</li><li>Google-Kairos Power: 500 MW advanced reactor partnership</li><li>Meta: 6.6 GW nuclear agreements</li></ul><p><strong>Related keywords:</strong> hyperscaler nuclear procurement, Microsoft Three Mile Island, Amazon nuclear investment, Google Kairos Power, data center nuclear PPA</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Thu, 05 Mar 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/43e06d0e/176ad665.mp3" length="1413969" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>170</itunes:duration>
      <itunes:summary>Microsoft, Amazon, Google, and Meta are contracting directly with nuclear generators—reshaping the entire nuclear industry.</itunes:summary>
      <itunes:subtitle>Microsoft, Amazon, Google, and Meta are contracting directly with nuclear generators—reshaping the entire nuclear industry.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Long-Duration Storage Changes the Renewables Equation</title>
      <itunes:episode>46</itunes:episode>
      <podcast:episode>46</podcast:episode>
      <itunes:title>Long-Duration Storage Changes the Renewables Equation</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">5bb65086-7476-4e2a-8fb0-30c4ee0bdfb2</guid>
      <link>https://share.transistor.fm/s/eda3f70f</link>
      <description>
        <![CDATA[<p>Long-duration energy storage is transforming the economics of renewable-powered data centers.</p><p><strong>Key topics covered:</strong></p><ul><li>Form Energy iron-air batteries: 100+ hour discharge duration</li><li>Chemistry basics: reversible rusting process stores and releases energy</li><li>West Virginia manufacturing facility under construction</li><li>Cost advantages from abundant iron vs. lithium/cobalt</li></ul><p><strong>Related keywords:</strong> long-duration energy storage, Form Energy, iron-air battery, 100-hour storage, renewable data centers, lithium alternative battery</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Long-duration energy storage is transforming the economics of renewable-powered data centers.</p><p><strong>Key topics covered:</strong></p><ul><li>Form Energy iron-air batteries: 100+ hour discharge duration</li><li>Chemistry basics: reversible rusting process stores and releases energy</li><li>West Virginia manufacturing facility under construction</li><li>Cost advantages from abundant iron vs. lithium/cobalt</li></ul><p><strong>Related keywords:</strong> long-duration energy storage, Form Energy, iron-air battery, 100-hour storage, renewable data centers, lithium alternative battery</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Wed, 04 Mar 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/eda3f70f/1ebf7b3f.mp3" length="1368422" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>164</itunes:duration>
      <itunes:summary>100-hour iron-air batteries are reaching commercial deployment, making renewable-powered data centers viable without fossil fuel backup.</itunes:summary>
      <itunes:subtitle>100-hour iron-air batteries are reaching commercial deployment, making renewable-powered data centers viable without fossil fuel backup.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Fusion's Long Bet on Data Center Power</title>
      <itunes:episode>45</itunes:episode>
      <podcast:episode>45</podcast:episode>
      <itunes:title>Fusion's Long Bet on Data Center Power</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">d9ecf4f0-a2c8-43a6-83c2-a6da3712935f</guid>
      <link>https://share.transistor.fm/s/68ae9b02</link>
      <description>
        <![CDATA[<p>Nuclear fusion represents the ultimate clean energy source for data centers—but when will it actually deliver power?</p><p><strong>Key topics covered:</strong></p><ul><li>Lawrence Livermore ignition milestone: net energy achieved December 2022</li><li>$10.6 billion private funding 2021-2025</li><li>Helion Energy-Microsoft PPA targeting 2028 delivery</li><li>Commonwealth Fusion Systems building 400 MW SPARC device</li></ul><p><strong>Related keywords:</strong> nuclear fusion data centers, Helion Energy Microsoft, Commonwealth Fusion Systems, SPARC reactor, fusion ignition, clean energy future</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Nuclear fusion represents the ultimate clean energy source for data centers—but when will it actually deliver power?</p><p><strong>Key topics covered:</strong></p><ul><li>Lawrence Livermore ignition milestone: net energy achieved December 2022</li><li>$10.6 billion private funding 2021-2025</li><li>Helion Energy-Microsoft PPA targeting 2028 delivery</li><li>Commonwealth Fusion Systems building 400 MW SPARC device</li></ul><p><strong>Related keywords:</strong> nuclear fusion data centers, Helion Energy Microsoft, Commonwealth Fusion Systems, SPARC reactor, fusion ignition, clean energy future</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Tue, 03 Mar 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/68ae9b02/ec4b29cf.mp3" length="1519499" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>183</itunes:duration>
      <itunes:summary>Nuclear fusion has been 20 years away for decades, but recent breakthroughs and $10.6 billion in private funding are compressing timelines.</itunes:summary>
      <itunes:subtitle>Nuclear fusion has been 20 years away for decades, but recent breakthroughs and $10.6 billion in private funding are compressing timelines.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Geothermal's Quiet Rise in Data Center Power</title>
      <itunes:episode>44</itunes:episode>
      <podcast:episode>44</podcast:episode>
      <itunes:title>Geothermal's Quiet Rise in Data Center Power</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">ecfae552-37aa-4abb-adb6-714381da06d4</guid>
      <link>https://share.transistor.fm/s/49baaf57</link>
      <description>
        <![CDATA[<p>Geothermal energy is emerging as a significant power source for data centers, with Google's Fervo Energy partnership leading commercial-scale deployment.</p><p><strong>Key topics covered:</strong></p><ul><li>Google-Fervo partnership: 115 MW enhanced geothermal in Nevada</li><li>Cape Station Utah: 100 MW in 2026, expanding to 2 GW</li><li>Horizontal drilling and fracking techniques from oil and gas</li><li>Ormat Technologies adding 150 MW for Google expansion</li></ul><p><strong>Related keywords:</strong> geothermal data centers, Fervo Energy, Google geothermal, enhanced geothermal systems, 24/7 carbon-free energy</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Geothermal energy is emerging as a significant power source for data centers, with Google's Fervo Energy partnership leading commercial-scale deployment.</p><p><strong>Key topics covered:</strong></p><ul><li>Google-Fervo partnership: 115 MW enhanced geothermal in Nevada</li><li>Cape Station Utah: 100 MW in 2026, expanding to 2 GW</li><li>Horizontal drilling and fracking techniques from oil and gas</li><li>Ormat Technologies adding 150 MW for Google expansion</li></ul><p><strong>Related keywords:</strong> geothermal data centers, Fervo Energy, Google geothermal, enhanced geothermal systems, 24/7 carbon-free energy</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Mon, 02 Mar 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/49baaf57/6f00400e.mp3" length="1457438" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>175</itunes:duration>
      <itunes:summary>While nuclear captures headlines, geothermal is quietly delivering 24/7 carbon-free baseload power to data centers today.</itunes:summary>
      <itunes:subtitle>While nuclear captures headlines, geothermal is quietly delivering 24/7 carbon-free baseload power to data centers today.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The NRC Bottleneck for Advanced Nuclear</title>
      <itunes:episode>43</itunes:episode>
      <podcast:episode>43</podcast:episode>
      <itunes:title>The NRC Bottleneck for Advanced Nuclear</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">7588e2ed-4ced-44d4-bd65-9828566622ea</guid>
      <link>https://share.transistor.fm/s/35e5d0b0</link>
      <description>
        <![CDATA[<p>The Nuclear Regulatory Commission's approval process determines when advanced reactors can actually power data centers.</p><p><strong>Key topics covered:</strong></p><ul><li>ADVANCE Act streamlining licensing for advanced reactors</li><li>Part 53 performance-based framework expected March 2026</li><li>NuScale 462 MW approval completed in under two years</li><li>TerraPower Natrium: first non-light water reactor approval in 50 years</li></ul><p><strong>Related keywords:</strong> NRC advanced reactor, ADVANCE Act nuclear, Part 53 framework, nuclear licensing timeline, TerraPower Natrium, SMR regulatory approval</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>The Nuclear Regulatory Commission's approval process determines when advanced reactors can actually power data centers.</p><p><strong>Key topics covered:</strong></p><ul><li>ADVANCE Act streamlining licensing for advanced reactors</li><li>Part 53 performance-based framework expected March 2026</li><li>NuScale 462 MW approval completed in under two years</li><li>TerraPower Natrium: first non-light water reactor approval in 50 years</li></ul><p><strong>Related keywords:</strong> NRC advanced reactor, ADVANCE Act nuclear, Part 53 framework, nuclear licensing timeline, TerraPower Natrium, SMR regulatory approval</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Fri, 27 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/35e5d0b0/e2c3dfc9.mp3" length="1507379" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>181</itunes:duration>
      <itunes:summary>The NRC is simultaneously the enabler and the constraint for advanced nuclear deployment.</itunes:summary>
      <itunes:subtitle>The NRC is simultaneously the enabler and the constraint for advanced nuclear deployment.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Westinghouse eVinci and the Truck-Delivered Reactor</title>
      <itunes:episode>42</itunes:episode>
      <podcast:episode>42</podcast:episode>
      <itunes:title>Westinghouse eVinci and the Truck-Delivered Reactor</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">286e1d20-e323-4ecc-bd6a-6c750f6a0391</guid>
      <link>https://share.transistor.fm/s/50e502f3</link>
      <description>
        <![CDATA[<p>The Westinghouse eVinci represents the most radical departure from traditional nuclear architecture—a reactor that fits on a truck and uses passive heat pipe cooling.</p><p><strong>Key topics covered:</strong></p><ul><li>5 MW output with 8-year fuel cycle and passive cooling</li><li>Heat pipe technology eliminating pumps and operator intervention</li><li>Smaller than a school bus, installable on 2-acre sites</li><li>Idaho National Laboratory testing scheduled for 2026</li></ul><p><strong>Related keywords:</strong> Westinghouse eVinci, microreactor, heat pipe reactor, passive nuclear cooling, truck-delivered reactor, DOE demonstration project</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>The Westinghouse eVinci represents the most radical departure from traditional nuclear architecture—a reactor that fits on a truck and uses passive heat pipe cooling.</p><p><strong>Key topics covered:</strong></p><ul><li>5 MW output with 8-year fuel cycle and passive cooling</li><li>Heat pipe technology eliminating pumps and operator intervention</li><li>Smaller than a school bus, installable on 2-acre sites</li><li>Idaho National Laboratory testing scheduled for 2026</li></ul><p><strong>Related keywords:</strong> Westinghouse eVinci, microreactor, heat pipe reactor, passive nuclear cooling, truck-delivered reactor, DOE demonstration project</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Thu, 26 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/50e502f3/98770b20.mp3" length="1374898" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>165</itunes:duration>
      <itunes:summary>Westinghouse's eVinci microreactor fits on a truck, runs 8 years without refueling, and requires no active cooling systems.</itunes:summary>
      <itunes:subtitle>Westinghouse's eVinci microreactor fits on a truck, runs 8 years without refueling, and requires no active cooling systems.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>NuScale's Pivot to AI Infrastructure</title>
      <itunes:episode>41</itunes:episode>
      <podcast:episode>41</podcast:episode>
      <itunes:title>NuScale's Pivot to AI Infrastructure</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">23ce55b7-c334-440a-bf69-42d0f488360a</guid>
      <link>https://share.transistor.fm/s/8ebac3ca</link>
      <description>
        <![CDATA[<p>NuScale Power is the only company with an NRC-certified small modular reactor design, and it's now targeting AI data centers as its primary market.</p><p><strong>Key topics covered:</strong></p><ul><li>NRC design certification advantage over competitors</li><li>VOYGR design: 77 MW per module, scalable to 924 MW</li><li>ENTRA1-TVA agreement: 6 GW, 72 modules</li><li>Standard Power: 2 GW for Ohio and Pennsylvania data centers by 2029</li></ul><p><strong>Related keywords:</strong> NuScale Power, SMR data centers, VOYGR reactor, ENTRA1 Energy, TVA nuclear agreement, NRC certified reactor</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>NuScale Power is the only company with an NRC-certified small modular reactor design, and it's now targeting AI data centers as its primary market.</p><p><strong>Key topics covered:</strong></p><ul><li>NRC design certification advantage over competitors</li><li>VOYGR design: 77 MW per module, scalable to 924 MW</li><li>ENTRA1-TVA agreement: 6 GW, 72 modules</li><li>Standard Power: 2 GW for Ohio and Pennsylvania data centers by 2029</li></ul><p><strong>Related keywords:</strong> NuScale Power, SMR data centers, VOYGR reactor, ENTRA1 Energy, TVA nuclear agreement, NRC certified reactor</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Wed, 25 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/8ebac3ca/cddb3880.mp3" length="1342909" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>161</itunes:duration>
      <itunes:summary>NuScale holds the only NRC-certified SMR design and is now pivoting hard toward AI data center applications.</itunes:summary>
      <itunes:subtitle>NuScale holds the only NRC-certified SMR design and is now pivoting hard toward AI data center applications.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Oklo's Bet on Data Center Nuclear</title>
      <itunes:episode>40</itunes:episode>
      <podcast:episode>40</podcast:episode>
      <itunes:title>Oklo's Bet on Data Center Nuclear</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">17a92929-8571-419b-a892-e87816318abc</guid>
      <link>https://share.transistor.fm/s/ade6f272</link>
      <description>
        <![CDATA[<p>Oklo is betting its future on powering AI data centers with micro reactors. This episode explores the company's Aurora powerhouse design, its strategic partnership with Meta, and the regulatory pathway to commercial operation.</p><p><strong>Key topics covered:</strong></p><ul><li>Aurora powerhouse: 15-50 MW units sized for data center campuses</li><li>Sam Altman-backed SPAC merger connecting nuclear to AI ecosystem</li><li>Sodium-cooled fast reactor design utilizing recycled nuclear fuel</li><li>Factory fabrication and truck transport logistics</li></ul><p><strong>Related keywords:</strong> Oklo nuclear, Aurora powerhouse, Meta data center nuclear, Sam Altman nuclear, sodium-cooled reactor, AI infrastructure power</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Oklo is betting its future on powering AI data centers with micro reactors. This episode explores the company's Aurora powerhouse design, its strategic partnership with Meta, and the regulatory pathway to commercial operation.</p><p><strong>Key topics covered:</strong></p><ul><li>Aurora powerhouse: 15-50 MW units sized for data center campuses</li><li>Sam Altman-backed SPAC merger connecting nuclear to AI ecosystem</li><li>Sodium-cooled fast reactor design utilizing recycled nuclear fuel</li><li>Factory fabrication and truck transport logistics</li></ul><p><strong>Related keywords:</strong> Oklo nuclear, Aurora powerhouse, Meta data center nuclear, Sam Altman nuclear, sodium-cooled reactor, AI infrastructure power</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Tue, 24 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/ade6f272/bf19673c.mp3" length="1436320" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>172</itunes:duration>
      <itunes:summary>Oklo has positioned itself as the micro reactor company purpose-built for AI infrastructure.</itunes:summary>
      <itunes:subtitle>Oklo has positioned itself as the micro reactor company purpose-built for AI infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Micro Reactors Are Coming Faster Than You Think</title>
      <itunes:episode>39</itunes:episode>
      <podcast:episode>39</podcast:episode>
      <itunes:title>Micro Reactors Are Coming Faster Than You Think</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">299b1d11-2d3b-4771-b5d0-bdb358efc57a</guid>
      <link>https://share.transistor.fm/s/d2d9bdb8</link>
      <description>
        <![CDATA[<p>The nuclear micro reactor is no longer theoretical—it's becoming a procurement decision for data center operators. This episode examines the leading micro reactor designs from Oklo and Westinghouse, their technical specifications, and the realistic timelines for commercial deployment.</p><p><strong>Key topics covered:</strong></p><ul><li>Oklo Aurora powerhouse: 15-50 MW units with sodium-cooled fast reactor design</li><li>Westinghouse eVinci: 5 MW output with 40-year lifespan and passive cooling</li><li>NRC licensing acceleration for advanced reactor designs</li><li>Factory fabrication economics vs. traditional nuclear construction</li><li>3-4 year deployment timelines vs. 5-7 year grid interconnection queues</li></ul><p><strong>Related keywords:</strong> micro reactors data centers, Oklo Aurora, Westinghouse eVinci, small modular reactors, NRC advanced reactor licensing, nuclear data center power</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>The nuclear micro reactor is no longer theoretical—it's becoming a procurement decision for data center operators. This episode examines the leading micro reactor designs from Oklo and Westinghouse, their technical specifications, and the realistic timelines for commercial deployment.</p><p><strong>Key topics covered:</strong></p><ul><li>Oklo Aurora powerhouse: 15-50 MW units with sodium-cooled fast reactor design</li><li>Westinghouse eVinci: 5 MW output with 40-year lifespan and passive cooling</li><li>NRC licensing acceleration for advanced reactor designs</li><li>Factory fabrication economics vs. traditional nuclear construction</li><li>3-4 year deployment timelines vs. 5-7 year grid interconnection queues</li></ul><p><strong>Related keywords:</strong> micro reactors data centers, Oklo Aurora, Westinghouse eVinci, small modular reactors, NRC advanced reactor licensing, nuclear data center power</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Mon, 23 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/d2d9bdb8/a5765517.mp3" length="1326829" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>159</itunes:duration>
      <itunes:summary>Micro reactors are moving from concept to procurement decision, with commercial deployments expected between 2027 and 2030.</itunes:summary>
      <itunes:subtitle>Micro reactors are moving from concept to procurement decision, with commercial deployments expected between 2027 and 2030.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Heat Recovery Turns Data Centers Into Utilities</title>
      <itunes:episode>38</itunes:episode>
      <podcast:episode>38</podcast:episode>
      <itunes:title>Heat Recovery Turns Data Centers Into Utilities</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">a4a3ca59-1fc7-4409-92be-99d4fd7bb115</guid>
      <link>https://share.transistor.fm/s/2636e276</link>
      <description>
        <![CDATA[<p>Data centers are becoming heating plants. Waste heat from AI workloads is being captured and redirected into district heating networks, transforming a byproduct into a revenue stream. This episode examines heat recovery projects across Europe and the business case for integrating heat recovery into new builds.</p><p><strong>Key topics covered:</strong></p><ul><li>Microsoft-Fortum partnership serving 250,000 people in Finland</li><li>Google's Hamina data center supplying 80% of district heating</li><li>AWS Dublin providing 92% of campus heating demand</li><li>Equinix Paris heating 1,000 homes from server waste heat</li><li>3-year payback on heat recovery infrastructure</li></ul><p><strong>Related keywords:</strong> data center heat recovery, district heating data centers, waste heat reuse, Microsoft Fortum Finland, ESG data centers, heat exchangers data centers</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Data centers are becoming heating plants. Waste heat from AI workloads is being captured and redirected into district heating networks, transforming a byproduct into a revenue stream. This episode examines heat recovery projects across Europe and the business case for integrating heat recovery into new builds.</p><p><strong>Key topics covered:</strong></p><ul><li>Microsoft-Fortum partnership serving 250,000 people in Finland</li><li>Google's Hamina data center supplying 80% of district heating</li><li>AWS Dublin providing 92% of campus heating demand</li><li>Equinix Paris heating 1,000 homes from server waste heat</li><li>3-year payback on heat recovery infrastructure</li></ul><p><strong>Related keywords:</strong> data center heat recovery, district heating data centers, waste heat reuse, Microsoft Fortum Finland, ESG data centers, heat exchangers data centers</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Fri, 20 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/2636e276/4c2ca4cc.mp3" length="1237177" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>147</itunes:duration>
      <itunes:summary>Microsoft and Fortum will supply 40% of district heating demand for 250,000 people in Finland using data center waste heat.</itunes:summary>
      <itunes:subtitle>Microsoft and Fortum will supply 40% of district heating demand for 250,000 people in Finland using data center waste heat.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Cabling Bottleneck You Haven't Heard About</title>
      <itunes:episode>37</itunes:episode>
      <podcast:episode>37</podcast:episode>
      <itunes:title>The Cabling Bottleneck You Haven't Heard About</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">080771dd-8c85-4fff-8c07-d41993cc9902</guid>
      <link>https://share.transistor.fm/s/e9324c60</link>
      <description>
        <![CDATA[<p>The networking stack for AI infrastructure is being upgraded simultaneously — from backend intra-cluster connections to frontend switch-to-switch links. Facilities that can't support high-speed networking become bottlenecks that throttle hardware worth tens of millions per rack. This episode examines the cabling requirements for AI-scale compute.</p><p><strong>Key topics covered:</strong></p><ul><li>Backend connections migrating to 1.6T and 3.2T speeds</li><li>NVIDIA Blackwell's 400-800 Gbps interconnect requirements</li><li>SN-MT and MMC connectors tripling fiber density</li><li>Fiber vs copper tradeoffs for AI networking</li><li>Supply chain constraints on high-speed cabling</li></ul><p><strong>Related keywords:</strong> data center cabling, high-speed interconnects, GPU networking, InfiniBand data centers, fiber optic AI, structured cabling AI</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>The networking stack for AI infrastructure is being upgraded simultaneously — from backend intra-cluster connections to frontend switch-to-switch links. Facilities that can't support high-speed networking become bottlenecks that throttle hardware worth tens of millions per rack. This episode examines the cabling requirements for AI-scale compute.</p><p><strong>Key topics covered:</strong></p><ul><li>Backend connections migrating to 1.6T and 3.2T speeds</li><li>NVIDIA Blackwell's 400-800 Gbps interconnect requirements</li><li>SN-MT and MMC connectors tripling fiber density</li><li>Fiber vs copper tradeoffs for AI networking</li><li>Supply chain constraints on high-speed cabling</li></ul><p><strong>Related keywords:</strong> data center cabling, high-speed interconnects, GPU networking, InfiniBand data centers, fiber optic AI, structured cabling AI</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Thu, 19 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/e9324c60/8473adc8.mp3" length="1319096" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>158</itunes:duration>
      <itunes:summary>GPU-to-GPU links are migrating to 1.6 terabits per second. The fastest GPU is limited by the speed of the cable connecting it to the next GPU.</itunes:summary>
      <itunes:subtitle>GPU-to-GPU links are migrating to 1.6 terabits per second. The fastest GPU is limited by the speed of the cable connecting it to the next GPU.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>15 Reactors Coming Online in 2026</title>
      <itunes:episode>36</itunes:episode>
      <podcast:episode>36</podcast:episode>
      <itunes:title>15 Reactors Coming Online in 2026</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">f42f7150-8f37-4458-a2a7-3104633eae99</guid>
      <link>https://share.transistor.fm/s/a1b7d09e</link>
      <description>
        <![CDATA[<p>Nuclear energy is becoming central to data center power strategy. With 15 reactors expected online in 2026 and hyperscalers signing multi-decade agreements, nuclear offers what solar and wind cannot: dispatchable carbon-free baseload power. This episode examines the nuclear pipeline and its implications for AI infrastructure.</p><p><strong>Key topics covered:</strong></p><ul><li>BloombergNEF projection of 15 reactors and 12 GW in 2026</li><li>Three Mile Island restart for Microsoft data centers</li><li>Amazon's $20 billion nuclear-powered data center investment</li><li>Google and Kairos Power 500 MW SMR partnership</li><li>Meta's 6.6 GW nuclear energy agreement</li></ul><p><strong>Related keywords:</strong> nuclear power data centers, SMR data centers, Three Mile Island Microsoft, Kairos Power Google, nuclear energy AI, reactor restart data centers</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Nuclear energy is becoming central to data center power strategy. With 15 reactors expected online in 2026 and hyperscalers signing multi-decade agreements, nuclear offers what solar and wind cannot: dispatchable carbon-free baseload power. This episode examines the nuclear pipeline and its implications for AI infrastructure.</p><p><strong>Key topics covered:</strong></p><ul><li>BloombergNEF projection of 15 reactors and 12 GW in 2026</li><li>Three Mile Island restart for Microsoft data centers</li><li>Amazon's $20 billion nuclear-powered data center investment</li><li>Google and Kairos Power 500 MW SMR partnership</li><li>Meta's 6.6 GW nuclear energy agreement</li></ul><p><strong>Related keywords:</strong> nuclear power data centers, SMR data centers, Three Mile Island Microsoft, Kairos Power Google, nuclear energy AI, reactor restart data centers</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Wed, 18 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/a1b7d09e/e850295f.mp3" length="2034002" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>247</itunes:duration>
      <itunes:summary>BloombergNEF expects 15 nuclear reactors to come online globally in 2026, adding 12 gigawatts. Tech giants are securing nuclear power for data centers.</itunes:summary>
      <itunes:subtitle>BloombergNEF expects 15 nuclear reactors to come online globally in 2026, adding 12 gigawatts. Tech giants are securing nuclear power for data centers.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Microgrids Are Becoming Primary Power</title>
      <itunes:episode>35</itunes:episode>
      <podcast:episode>35</podcast:episode>
      <itunes:title>Microgrids Are Becoming Primary Power</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">1b3b16e2-35a0-4a15-a7fc-4a42eb73d5b6</guid>
      <link>https://share.transistor.fm/s/1809f6f7</link>
      <description>
        <![CDATA[<p>Microgrids are shifting from supplemental backup to primary power strategy. With grid connection delays extending up to seven years in some markets, operators are building their own power infrastructure. This episode examines the architectures, economics, and strategic implications of microgrid adoption.</p><p><strong>Key topics covered:</strong></p><ul><li>Grid constraints driving on-site power adoption</li><li>Hybrid AC/DC microgrid architectures</li><li>One-third of data centers targeting 100% on-site power by 2030</li><li>Fuel flexibility and dispatchable power for AI workloads</li><li>AI-driven energy management systems</li></ul><p><strong>Related keywords:</strong> data center microgrids, on-site power generation, grid independence, hybrid AC DC microgrid, dispatchable power AI, energy management systems</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Microgrids are shifting from supplemental backup to primary power strategy. With grid connection delays extending up to seven years in some markets, operators are building their own power infrastructure. This episode examines the architectures, economics, and strategic implications of microgrid adoption.</p><p><strong>Key topics covered:</strong></p><ul><li>Grid constraints driving on-site power adoption</li><li>Hybrid AC/DC microgrid architectures</li><li>One-third of data centers targeting 100% on-site power by 2030</li><li>Fuel flexibility and dispatchable power for AI workloads</li><li>AI-driven energy management systems</li></ul><p><strong>Related keywords:</strong> data center microgrids, on-site power generation, grid independence, hybrid AC DC microgrid, dispatchable power AI, energy management systems</p><p>Brought to you by Spring Street Management Group.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Tue, 17 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/1809f6f7/41b0873c.mp3" length="1285232" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>153</itunes:duration>
      <itunes:summary>92% of data center decision-makers see grid constraints as a major obstacle. One-third of data centers may rely on 100% on-site power by 2030.</itunes:summary>
      <itunes:subtitle>92% of data center decision-makers see grid constraints as a major obstacle. One-third of data centers may rely on 100% on-site power by 2030.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Liquid Cooling Hits 40% of AI Data Centers</title>
      <itunes:episode>34</itunes:episode>
      <podcast:episode>34</podcast:episode>
      <itunes:title>Liquid Cooling Hits 40% of AI Data Centers</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">deed5198-77c8-4376-acc7-9fe54a238122</guid>
      <link>https://share.transistor.fm/s/40c7710f</link>
      <description>
        <![CDATA[Liquid cooling is no longer optional for AI infrastructure — it's the baseline requirement. The data center cooling market is projected to reach $31 billion by 2034. This episode examines the technologies, infrastructure implications, and heat recovery opportunities.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>40% of AI data centers deploying liquid cooling in 2026</li>
<li>Cooling market growing from $11.65B to $31B by 2034</li>
<li>Direct-to-chip, immersion, and CDU-based systems</li>
<li>AI-driven thermal management systems</li>
<li>Microsoft-Fortum heat recovery for district heating</li>
</ul>

<p><strong>Related keywords:</strong> liquid cooling data centers, immersion cooling AI, direct-to-chip cooling, data center thermal management, heat recovery data centers</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Liquid cooling is no longer optional for AI infrastructure — it's the baseline requirement. The data center cooling market is projected to reach $31 billion by 2034. This episode examines the technologies, infrastructure implications, and heat recovery opportunities.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>40% of AI data centers deploying liquid cooling in 2026</li>
<li>Cooling market growing from $11.65B to $31B by 2034</li>
<li>Direct-to-chip, immersion, and CDU-based systems</li>
<li>AI-driven thermal management systems</li>
<li>Microsoft-Fortum heat recovery for district heating</li>
</ul>

<p><strong>Related keywords:</strong> liquid cooling data centers, immersion cooling AI, direct-to-chip cooling, data center thermal management, heat recovery data centers</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Mon, 16 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/40c7710f/d0655429.mp3" length="1253263" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>150</itunes:duration>
      <itunes:summary>Liquid cooling is expected in 40% of AI data centers in 2026. At 100+ kW per rack, air cooling cannot remove heat fast enough.</itunes:summary>
      <itunes:subtitle>Liquid cooling is expected in 40% of AI data centers in 2026. At 100+ kW per rack, air cooling cannot remove heat fast enough.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Permitting Problem Is Getting Worse</title>
      <itunes:episode>33</itunes:episode>
      <podcast:episode>33</podcast:episode>
      <itunes:title>The Permitting Problem Is Getting Worse</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">fdb7afdb-ce66-40be-99bc-3871d4e7ad0f</guid>
      <link>https://share.transistor.fm/s/060d0227</link>
      <description>
        <![CDATA[Data center permitting timelines have tripled in legacy markets, and community opposition has blocked billions in projects. This episode examines the permitting landscape, regulatory fragmentation, and strategies for navigating local politics.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Permitting extending from 6-12 months to 2-3 years</li>
<li>New York moratorium bill for 20+ MW data centers</li>
<li>Community opposition and misconceptions</li>
<li>State and local regulatory fragmentation</li>
<li>Brownfield guidance and federal expediting efforts</li>
</ul>

<p><strong>Related keywords:</strong> data center permitting, data center moratorium, community opposition data centers, data center zoning, permitting delays AI infrastructure</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Data center permitting timelines have tripled in legacy markets, and community opposition has blocked billions in projects. This episode examines the permitting landscape, regulatory fragmentation, and strategies for navigating local politics.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Permitting extending from 6-12 months to 2-3 years</li>
<li>New York moratorium bill for 20+ MW data centers</li>
<li>Community opposition and misconceptions</li>
<li>State and local regulatory fragmentation</li>
<li>Brownfield guidance and federal expediting efforts</li>
</ul>

<p><strong>Related keywords:</strong> data center permitting, data center moratorium, community opposition data centers, data center zoning, permitting delays AI infrastructure</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Fri, 13 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/060d0227/ffee530d.mp3" length="1214181" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>145</itunes:duration>
      <itunes:summary>Permitting timelines have increased from 6-12 months to 2-3 years in legacy markets. New York may impose a moratorium on large data center permits.</itunes:summary>
      <itunes:subtitle>Permitting timelines have increased from 6-12 months to 2-3 years in legacy markets. New York may impose a moratorium on large data center permits.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Meta Compute and the 5-Gigawatt Campus</title>
      <itunes:episode>32</itunes:episode>
      <podcast:episode>32</podcast:episode>
      <itunes:title>Meta Compute and the 5-Gigawatt Campus</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">a1535f60-c02b-4e14-8195-e2e00cf1022e</guid>
      <link>https://share.transistor.fm/s/20ab4400</link>
      <description>
        <![CDATA[Meta has created Meta Compute to centralize and accelerate its AI infrastructure buildout. With Prometheus in Ohio and Hyperion scaling to 5 gigawatts in Louisiana, Meta is building at scales that rival utility infrastructure projects. This episode examines Meta's compute strategy and its implications.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Meta Compute division reporting directly to Zuckerberg</li>
<li>Prometheus supercluster activating in Ohio 2026</li>
<li>Hyperion campus scaling to 5 GW in Louisiana</li>
<li>$135 billion AI investment potential in 2026</li>
<li>Llama 4 and distributed inference requirements</li>
</ul>

<p><strong>Related keywords:</strong> Meta Compute, Meta data centers, Hyperion Louisiana, Prometheus Ohio, Meta AI infrastructure, Zuckerberg compute</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Meta has created Meta Compute to centralize and accelerate its AI infrastructure buildout. With Prometheus in Ohio and Hyperion scaling to 5 gigawatts in Louisiana, Meta is building at scales that rival utility infrastructure projects. This episode examines Meta's compute strategy and its implications.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Meta Compute division reporting directly to Zuckerberg</li>
<li>Prometheus supercluster activating in Ohio 2026</li>
<li>Hyperion campus scaling to 5 GW in Louisiana</li>
<li>$135 billion AI investment potential in 2026</li>
<li>Llama 4 and distributed inference requirements</li>
</ul>

<p><strong>Related keywords:</strong> Meta Compute, Meta data centers, Hyperion Louisiana, Prometheus Ohio, Meta AI infrastructure, Zuckerberg compute</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Thu, 12 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/20ab4400/6336a1e5.mp3" length="2153125" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>262</itunes:duration>
      <itunes:summary>Meta created Meta Compute to build computing capacity faster than anyone else. Hyperion in Louisiana will scale to 5 gigawatts — equivalent to 5 nuclear reactors.</itunes:summary>
      <itunes:subtitle>Meta created Meta Compute to build computing capacity faster than anyone else. Hyperion in Louisiana will scale to 5 gigawatts — equivalent to 5 nuclear reactors.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Fuel Cells Just Became a $5 Billion Business</title>
      <itunes:episode>31</itunes:episode>
      <podcast:episode>31</podcast:episode>
      <itunes:title>Fuel Cells Just Became a $5 Billion Business</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">8b47ebcd-0f45-49aa-9a33-a55d0802c88b</guid>
      <link>https://share.transistor.fm/s/981fd17b</link>
      <description>
        <![CDATA[Fuel cell adoption has moved from small-scale pilots to multi-billion-dollar primary power deployments for AI data centers. This episode examines the technology, business models, and market shift driving fuel cells as a time-to-power solution.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Bloom Energy's $5 billion Brookfield partnership</li>
<li>$2.65 billion American Electric Power agreement</li>
<li>Solid oxide fuel cell efficiency advantages</li>
<li>ECL's 1 GW hydrogen-powered AI Factory in Texas</li>
<li>Hydrogen as storage medium for 24/7 carbon-free energy</li>
</ul>

<p><strong>Related keywords:</strong> fuel cells data centers, Bloom Energy Brookfield, solid oxide fuel cells, hydrogen data centers, on-site power generation AI</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Fuel cell adoption has moved from small-scale pilots to multi-billion-dollar primary power deployments for AI data centers. This episode examines the technology, business models, and market shift driving fuel cells as a time-to-power solution.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Bloom Energy's $5 billion Brookfield partnership</li>
<li>$2.65 billion American Electric Power agreement</li>
<li>Solid oxide fuel cell efficiency advantages</li>
<li>ECL's 1 GW hydrogen-powered AI Factory in Texas</li>
<li>Hydrogen as storage medium for 24/7 carbon-free energy</li>
</ul>

<p><strong>Related keywords:</strong> fuel cells data centers, Bloom Energy Brookfield, solid oxide fuel cells, hydrogen data centers, on-site power generation AI</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Wed, 11 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/981fd17b/152f71ec.mp3" length="1308227" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>156</itunes:duration>
      <itunes:summary>Bloom Energy secured a $5 billion partnership with Brookfield for solid oxide fuel cells. Fuel cells have shifted from pilots to primary power deployments.</itunes:summary>
      <itunes:subtitle>Bloom Energy secured a $5 billion partnership with Brookfield for solid oxide fuel cells. Fuel cells have shifted from pilots to primary power deployments.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Copper Crisis No One Is Pricing In</title>
      <itunes:episode>30</itunes:episode>
      <podcast:episode>30</podcast:episode>
      <itunes:title>The Copper Crisis No One Is Pricing In</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">dd735ec8-0285-4474-9cea-0d1693d6a559</guid>
      <link>https://share.transistor.fm/s/b3772453</link>
      <description>
        <![CDATA[AI-driven data centers require about four times more copper than traditional facilities. With a global deficit forecast and prices projected to hit $12,500-$13,000 per ton, copper is becoming a critical constraint on AI infrastructure. This episode examines the copper supply chain and its implications for investors.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>AI data centers requiring 4x copper per facility</li>
<li>Global refined copper deficit of 150,000 metric tons in 2026</li>
<li>JPMorgan and UBS copper price projections</li>
<li>Strategic stockpiling and blanket agreements</li>
<li>Aluminum as partial substitute for busway systems</li>
</ul>

<p><strong>Related keywords:</strong> copper supply data centers, copper deficit AI, copper price forecast, data center materials, copper demand AI infrastructure</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[AI-driven data centers require about four times more copper than traditional facilities. With a global deficit forecast and prices projected to hit $12,500-$13,000 per ton, copper is becoming a critical constraint on AI infrastructure. This episode examines the copper supply chain and its implications for investors.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>AI data centers requiring 4x copper per facility</li>
<li>Global refined copper deficit of 150,000 metric tons in 2026</li>
<li>JPMorgan and UBS copper price projections</li>
<li>Strategic stockpiling and blanket agreements</li>
<li>Aluminum as partial substitute for busway systems</li>
</ul>

<p><strong>Related keywords:</strong> copper supply data centers, copper deficit AI, copper price forecast, data center materials, copper demand AI infrastructure</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Tue, 10 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/b3772453/ea42ccc3.mp3" length="1346673" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>161</itunes:duration>
      <itunes:summary>AI data centers require 4x more copper than traditional facilities. The market is pricing AI growth but may not be pricing the copper required to build it.</itunes:summary>
      <itunes:subtitle>AI data centers require 4x more copper than traditional facilities. The market is pricing AI growth but may not be pricing the copper required to build it.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The $690 Billion Capex Arms Race</title>
      <itunes:episode>29</itunes:episode>
      <podcast:episode>29</podcast:episode>
      <itunes:title>The $690 Billion Capex Arms Race</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">d73fb1ba-fc77-4773-8f07-75995ae99ee1</guid>
      <link>https://share.transistor.fm/s/3c6deced</link>
      <description>
        <![CDATA[Five tech giants are deploying nearly $700 billion in a single year to build computing infrastructure — more than the GDP of most countries. This episode examines the capital deployment driving the AI buildout and its downstream effects on utilities, construction, and materials.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Amazon leading with $200 billion capex</li>
<li>Alphabet nearly doubling to $175-185 billion</li>
<li>Meta Compute targeting $115-135 billion</li>
<li>Microsoft Azure expansion at $110-145 billion</li>
<li>Supply chain strain from unprecedented capital deployment</li>
</ul>

<p><strong>Related keywords:</strong> hyperscaler capex, AI infrastructure spending, Amazon AWS data centers, Meta Compute, Microsoft Azure expansion, tech capital expenditure 2026</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Five tech giants are deploying nearly $700 billion in a single year to build computing infrastructure — more than the GDP of most countries. This episode examines the capital deployment driving the AI buildout and its downstream effects on utilities, construction, and materials.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Amazon leading with $200 billion capex</li>
<li>Alphabet nearly doubling to $175-185 billion</li>
<li>Meta Compute targeting $115-135 billion</li>
<li>Microsoft Azure expansion at $110-145 billion</li>
<li>Supply chain strain from unprecedented capital deployment</li>
</ul>

<p><strong>Related keywords:</strong> hyperscaler capex, AI infrastructure spending, Amazon AWS data centers, Meta Compute, Microsoft Azure expansion, tech capital expenditure 2026</p>

<p>Brought to you by Spring Street Management Group.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Mon, 09 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/3c6deced/13765039.mp3" length="1410615" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>169</itunes:duration>
      <itunes:summary>Amazon, Alphabet, Meta, Microsoft, and Oracle are projected to spend $660-$690 billion on capex in 2026 — nearly double 2025 levels.</itunes:summary>
      <itunes:subtitle>Amazon, Alphabet, Meta, Microsoft, and Oracle are projected to spend $660-$690 billion on capex in 2026 — nearly double 2025 levels.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>PJM Just Authorized Redirecting Data Center Power to Homes</title>
      <itunes:episode>28</itunes:episode>
      <podcast:episode>28</podcast:episode>
      <itunes:title>PJM Just Authorized Redirecting Data Center Power to Homes</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">976bb889-b99e-4024-acde-3e1263717755</guid>
      <link>https://share.transistor.fm/s/6231d4a2</link>
      <description>
        <![CDATA[PJM—the grid operator for 13 states—now has authorization to redirect power from data centers to households during grid emergencies. This episode examines what this policy shift means for data center operators and the evolving relationship between AI infrastructure and grid reliability.<p><strong>Key topics covered:</strong></p><ul><li>PJM's new authority to require data centers to switch to backup generators</li><li>The blurring line between backup power and dispatchable generation</li><li>Large load tariffs and demand response provisions for data centers</li><li>Google and Fervo Energy's clean transition tariff with NV Energy</li><li>Implications for backup generator maintenance, fuel contracts, and emissions permits</li></ul><p><strong>Related keywords:</strong> PJM data centers, grid reliability, demand response, backup generators, large load tariff, data center curtailment, Google Fervo geothermal</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[PJM—the grid operator for 13 states—now has authorization to redirect power from data centers to households during grid emergencies. This episode examines what this policy shift means for data center operators and the evolving relationship between AI infrastructure and grid reliability.<p><strong>Key topics covered:</strong></p><ul><li>PJM's new authority to require data centers to switch to backup generators</li><li>The blurring line between backup power and dispatchable generation</li><li>Large load tariffs and demand response provisions for data centers</li><li>Google and Fervo Energy's clean transition tariff with NV Energy</li><li>Implications for backup generator maintenance, fuel contracts, and emissions permits</li></ul><p><strong>Related keywords:</strong> PJM data centers, grid reliability, demand response, backup generators, large load tariff, data center curtailment, Google Fervo geothermal</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </content:encoded>
      <pubDate>Fri, 06 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/6231d4a2/41808c26.mp3" length="1304479" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>156</itunes:duration>
      <itunes:summary>How grid operators are treating data center backup generation as a dispatchable resource during emergencies.</itunes:summary>
      <itunes:subtitle>How grid operators are treating data center backup generation as a dispatchable resource during emergencies.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Blackwell Changes Everything About Data Center Design</title>
      <itunes:episode>27</itunes:episode>
      <podcast:episode>27</podcast:episode>
      <itunes:title>Blackwell Changes Everything About Data Center Design</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">5af36602-3783-45d5-825c-8c847d2db5cf</guid>
      <link>https://share.transistor.fm/s/5c08bf30</link>
      <description>
        <![CDATA[NVIDIA's Blackwell GPUs demand 10-30x the power density of traditional data centers, rendering most existing facilities obsolete for AI workloads. This episode breaks down the infrastructure overhaul required to support next-generation AI hardware.<p><strong>Key topics covered:</strong></p><ul><li>Blackwell rack power requirements: 60-140 kilowatts vs. traditional 4-6 kilowatts</li><li>Why fewer than 5% of data centers can support 50+ kW racks</li><li>Liquid cooling as a hard requirement, not an option</li><li>Electrical infrastructure overhaul: transformers, PDUs, bus bars, and wiring</li><li>400-800 Gbps networking with Quantum-X800 InfiniBand</li></ul><p><strong>Related keywords:</strong> NVIDIA Blackwell, data center power density, liquid cooling AI, NVL72, high-density computing, AI infrastructure design, colocation AI</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[NVIDIA's Blackwell GPUs demand 10-30x the power density of traditional data centers, rendering most existing facilities obsolete for AI workloads. This episode breaks down the infrastructure overhaul required to support next-generation AI hardware.<p><strong>Key topics covered:</strong></p><ul><li>Blackwell rack power requirements: 60-140 kilowatts vs. traditional 4-6 kilowatts</li><li>Why fewer than 5% of data centers can support 50+ kW racks</li><li>Liquid cooling as a hard requirement, not an option</li><li>Electrical infrastructure overhaul: transformers, PDUs, bus bars, and wiring</li><li>400-800 Gbps networking with Quantum-X800 InfiniBand</li></ul><p><strong>Related keywords:</strong> NVIDIA Blackwell, data center power density, liquid cooling AI, NVL72, high-density computing, AI infrastructure design, colocation AI</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </content:encoded>
      <pubDate>Thu, 05 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/5c08bf30/cad3b290.mp3" length="1280023" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>153</itunes:duration>
      <itunes:summary>How NVIDIA's Blackwell architecture is forcing a complete rethink of data center infrastructure requirements.</itunes:summary>
      <itunes:subtitle>How NVIDIA's Blackwell architecture is forcing a complete rethink of data center infrastructure requirements.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why Texas Will Have 40 Gigawatts of Data Centers by 2028</title>
      <itunes:episode>26</itunes:episode>
      <podcast:episode>26</podcast:episode>
      <itunes:title>Why Texas Will Have 40 Gigawatts of Data Centers by 2028</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">96747205-95f5-41d8-96a6-c26575f5641c</guid>
      <link>https://share.transistor.fm/s/986be7dc</link>
      <description>
        <![CDATA[Texas is on track to reach 40 gigawatts of data center capacity by 2028—a 142% increase from current levels and more than the total capacity of most countries. This episode examines why the Lone Star State is winning the race for AI infrastructure investment.<p><strong>Key topics covered:</strong></p><ul><li>ERCOT's regulatory advantages over FERC-jurisdictional regions</li><li>OpenAI's 1.2GW Stargate data center in Milam County with SB Energy</li><li>Galaxy's 1.6GW Helios campus expansion in West Texas</li><li>West Texas oil patch transformation into a data center corridor</li><li>Co-location of gas generation with compute to eliminate grid dependencies</li></ul><p><strong>Related keywords:</strong> Texas data centers, ERCOT, Stargate Texas, SB Energy OpenAI, Galaxy Helios, West Texas AI infrastructure, data center permitting</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Texas is on track to reach 40 gigawatts of data center capacity by 2028—a 142% increase from current levels and more than the total capacity of most countries. This episode examines why the Lone Star State is winning the race for AI infrastructure investment.<p><strong>Key topics covered:</strong></p><ul><li>ERCOT's regulatory advantages over FERC-jurisdictional regions</li><li>OpenAI's 1.2GW Stargate data center in Milam County with SB Energy</li><li>Galaxy's 1.6GW Helios campus expansion in West Texas</li><li>West Texas oil patch transformation into a data center corridor</li><li>Co-location of gas generation with compute to eliminate grid dependencies</li></ul><p><strong>Related keywords:</strong> Texas data centers, ERCOT, Stargate Texas, SB Energy OpenAI, Galaxy Helios, West Texas AI infrastructure, data center permitting</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </content:encoded>
      <pubDate>Wed, 04 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/986be7dc/51346c35.mp3" length="1337705" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>160</itunes:duration>
      <itunes:summary>How ERCOT independence, abundant power, and streamlined permitting are making Texas the dominant destination for AI infrastructure.</itunes:summary>
      <itunes:subtitle>How ERCOT independence, abundant power, and streamlined permitting are making Texas the dominant destination for AI infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Labor Crisis Threatening the AI Buildout</title>
      <itunes:episode>25</itunes:episode>
      <podcast:episode>25</podcast:episode>
      <itunes:title>The Labor Crisis Threatening the AI Buildout</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">a244b616-b974-4c82-b860-f9ede8d43b8d</guid>
      <link>https://share.transistor.fm/s/da57ba32</link>
      <description>
        <![CDATA[The AI infrastructure buildout faces a constraint that capital alone cannot solve: a severe shortage of skilled construction labor. This episode explores how 82% of construction firms report difficulty filling positions, and why projects are being postponed or canceled despite secured funding and permits.<p><strong>Key topics covered:</strong></p><ul><li>Peak crew sizes jumping from 750 to 4,000-5,000 workers per site</li><li>82% of firms struggling to fill hourly craft positions</li><li>60% of projects postponed or canceled due to labor constraints</li><li>Worker migration from Arizona to Dallas as markets shift</li><li>Competition for electrical workers from EVs, heat pumps, and grid upgrades</li></ul><p><strong>Related keywords:</strong> data center construction, skilled labor shortage, electrical workforce, construction megaprojects, AI infrastructure labor, DataBank Red Oak</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[The AI infrastructure buildout faces a constraint that capital alone cannot solve: a severe shortage of skilled construction labor. This episode explores how 82% of construction firms report difficulty filling positions, and why projects are being postponed or canceled despite secured funding and permits.<p><strong>Key topics covered:</strong></p><ul><li>Peak crew sizes jumping from 750 to 4,000-5,000 workers per site</li><li>82% of firms struggling to fill hourly craft positions</li><li>60% of projects postponed or canceled due to labor constraints</li><li>Worker migration from Arizona to Dallas as markets shift</li><li>Competition for electrical workers from EVs, heat pumps, and grid upgrades</li></ul><p><strong>Related keywords:</strong> data center construction, skilled labor shortage, electrical workforce, construction megaprojects, AI infrastructure labor, DataBank Red Oak</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </content:encoded>
      <pubDate>Tue, 03 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/da57ba32/98d2a13e.mp3" length="1300494" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>155</itunes:duration>
      <itunes:summary>Why the skilled labor shortage in data center construction is becoming a critical bottleneck for AI infrastructure.</itunes:summary>
      <itunes:subtitle>Why the skilled labor shortage in data center construction is becoming a critical bottleneck for AI infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Natural Gas Is Winning the AI Power Race</title>
      <itunes:episode>24</itunes:episode>
      <podcast:episode>24</podcast:episode>
      <itunes:title>Natural Gas Is Winning the AI Power Race</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">c05dc06a-c016-4832-8f34-ab31b0c35c50</guid>
      <link>https://share.transistor.fm/s/ae8851bb</link>
      <description>
        <![CDATA[With interconnection queues stretching 4-7 years, AI data center developers are turning to natural gas generation at unprecedented scale. This episode examines how 48 gigawatts of proposed capacity—a third of all U.S. projects—are now designed to bypass the grid entirely using behind-the-meter gas turbines.<p><strong>Key topics covered:</strong></p><ul><li>The 24x increase in behind-the-meter gas generation projects in two years</li><li>Caterpillar's 2GW deal for the Monarch Compute Campus in West Virginia</li><li>Pennsylvania's 4GW+ Homer City gas plant for AI data centers</li><li>Energy Transfer's 6 billion cubic feet/day pipeline expansion for AI infrastructure</li><li>The tension between decarbonization goals and immediate power needs</li></ul><p><strong>Related keywords:</strong> natural gas data centers, behind-the-meter generation, AI power infrastructure, gas turbines, interconnection queues, Caterpillar data center, Energy Transfer AI</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[With interconnection queues stretching 4-7 years, AI data center developers are turning to natural gas generation at unprecedented scale. This episode examines how 48 gigawatts of proposed capacity—a third of all U.S. projects—are now designed to bypass the grid entirely using behind-the-meter gas turbines.<p><strong>Key topics covered:</strong></p><ul><li>The 24x increase in behind-the-meter gas generation projects in two years</li><li>Caterpillar's 2GW deal for the Monarch Compute Campus in West Virginia</li><li>Pennsylvania's 4GW+ Homer City gas plant for AI data centers</li><li>Energy Transfer's 6 billion cubic feet/day pipeline expansion for AI infrastructure</li><li>The tension between decarbonization goals and immediate power needs</li></ul><p><strong>Related keywords:</strong> natural gas data centers, behind-the-meter generation, AI power infrastructure, gas turbines, interconnection queues, Caterpillar data center, Energy Transfer AI</p><p><em>The Power Allocation is brought to you by Spring Street Management Group.</em></p>]]&gt;]]>
      </content:encoded>
      <pubDate>Mon, 02 Feb 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/ae8851bb/1feaf3ad.mp3" length="1306969" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>156</itunes:duration>
      <itunes:summary>How behind-the-meter natural gas generation became the dominant solution for AI data center power needs.</itunes:summary>
      <itunes:subtitle>How behind-the-meter natural gas generation became the dominant solution for AI data center power needs.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Europe Is Losing the Data Center Race</title>
      <itunes:episode>23</itunes:episode>
      <podcast:episode>23</podcast:episode>
      <itunes:title>Europe Is Losing the Data Center Race</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">dd33cf6e-f37c-47da-8ff4-d497b23fff19</guid>
      <link>https://share.transistor.fm/s/31eb96ea</link>
      <description>
        <![CDATA[<p>Traditional European data center hubs face power shortages, land constraints, and regulatory complexity that are pushing new AI development elsewhere. The infrastructure Europe built for the cloud era may not serve the AI era.</p><p><strong>Key topics covered:</strong></p><ul><li>Dublin, Amsterdam, Frankfurt: structural constraints on new development</li><li>Google's €5.5 billion Germany investment struggling to find power</li><li>Asia-Pacific growth: 32 GW to 57 GW by 2030</li><li>EU sustainability regulations and compliance friction</li><li>Nordic exceptions: Sweden, Finland, Norway as alternative hubs</li><li>Capital flowing to Texas, Mississippi, Abu Dhabi, Singapore</li></ul><p><strong>Related keywords:</strong> Europe data centers, Dublin grid constraints, Frankfurt data center, EU data center regulations, Nordic data centers, AI infrastructure Europe, data center site selection, hyperscale capacity</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Traditional European data center hubs face power shortages, land constraints, and regulatory complexity that are pushing new AI development elsewhere. The infrastructure Europe built for the cloud era may not serve the AI era.</p><p><strong>Key topics covered:</strong></p><ul><li>Dublin, Amsterdam, Frankfurt: structural constraints on new development</li><li>Google's €5.5 billion Germany investment struggling to find power</li><li>Asia-Pacific growth: 32 GW to 57 GW by 2030</li><li>EU sustainability regulations and compliance friction</li><li>Nordic exceptions: Sweden, Finland, Norway as alternative hubs</li><li>Capital flowing to Texas, Mississippi, Abu Dhabi, Singapore</li></ul><p><strong>Related keywords:</strong> Europe data centers, Dublin grid constraints, Frankfurt data center, EU data center regulations, Nordic data centers, AI infrastructure Europe, data center site selection, hyperscale capacity</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </content:encoded>
      <pubDate>Fri, 30 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/31eb96ea/e01bbadc.mp3" length="1927635" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>234</itunes:duration>
      <itunes:summary>Dublin is closed. Amsterdam is constrained. Frankfurt faces limits. AI demand is redrawing Europe's data center map — and not in Europe's favor.</itunes:summary>
      <itunes:subtitle>Dublin is closed. Amsterdam is constrained. Frankfurt faces limits. AI demand is redrawing Europe's data center map — and not in Europe's favor.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why SMRs Won't Save Us Until 2030</title>
      <itunes:episode>22</itunes:episode>
      <podcast:episode>22</podcast:episode>
      <itunes:title>Why SMRs Won't Save Us Until 2030</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">62bf26fe-6c2a-4baf-acc3-d76a316e8425</guid>
      <link>https://share.transistor.fm/s/e87ae5dd</link>
      <description>
        <![CDATA[<p>Small modular reactors have become the energy solution every AI company wants to talk about. But the timeline reality is stark: the earliest SMR-powered data centers won't come online until late 2027 at the absolute earliest, with more realistic estimates pointing to 2030 or beyond.</p><p><strong>Key topics covered:</strong></p><ul><li>Hyperscaler nuclear deals: Meta/Oklo, Amazon/X-energy, Microsoft/Constellation, Google/Kairos</li><li>Actual SMR timelines: Oklo Aurora (2028), TerraPower Natrium (2030), Kairos/Google (2030)</li><li>NuScale's cancelled Utah project and lessons learned</li><li>First-of-a-kind nuclear: regulatory review, financing, and execution risk</li><li>The timing mismatch between AI demand growth and SMR availability</li><li>What fills the gap: natural gas, grid power, and existing nuclear restarts</li></ul><p><strong>Related keywords:</strong> small modular reactors, SMR data centers, Oklo, NuScale, TerraPower, Kairos Power, nuclear energy AI, Three Mile Island restart, baseload generation</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Small modular reactors have become the energy solution every AI company wants to talk about. But the timeline reality is stark: the earliest SMR-powered data centers won't come online until late 2027 at the absolute earliest, with more realistic estimates pointing to 2030 or beyond.</p><p><strong>Key topics covered:</strong></p><ul><li>Hyperscaler nuclear deals: Meta/Oklo, Amazon/X-energy, Microsoft/Constellation, Google/Kairos</li><li>Actual SMR timelines: Oklo Aurora (2028), TerraPower Natrium (2030), Kairos/Google (2030)</li><li>NuScale's cancelled Utah project and lessons learned</li><li>First-of-a-kind nuclear: regulatory review, financing, and execution risk</li><li>The timing mismatch between AI demand growth and SMR availability</li><li>What fills the gap: natural gas, grid power, and existing nuclear restarts</li></ul><p><strong>Related keywords:</strong> small modular reactors, SMR data centers, Oklo, NuScale, TerraPower, Kairos Power, nuclear energy AI, Three Mile Island restart, baseload generation</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </content:encoded>
      <pubDate>Thu, 29 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/e87ae5dd/5f07d932.mp3" length="2008506" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>244</itunes:duration>
      <itunes:summary>Every hyperscaler is betting on small modular reactors. But the earliest SMR-powered data centers won't come online until 2030. Here's the timeline reality.</itunes:summary>
      <itunes:subtitle>Every hyperscaler is betting on small modular reactors. But the earliest SMR-powered data centers won't come online until 2030. Here's the timeline reality.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Middle East Enters the AI Infrastructure Race</title>
      <itunes:episode>21</itunes:episode>
      <podcast:episode>21</podcast:episode>
      <itunes:title>The Middle East Enters the AI Infrastructure Race</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">43f3c6b0-4b43-466e-a50c-5bceeeb92936</guid>
      <link>https://share.transistor.fm/s/96cc2e0a</link>
      <description>
        <![CDATA[<p>AI data center investments across the Gulf Cooperation Council countries are projected to exceed $5 billion in 2026. The Middle East has entered the AI infrastructure race with structural advantages that constrained Western markets cannot match.</p><p><strong>Key topics covered:</strong></p><ul><li>Saudi Arabia's Humain: $1.2 billion for 250 megawatts of AI capacity</li><li>OpenAI Stargate UAE: 200MW first phase of a 1 gigawatt cluster</li><li>Sovereign wealth funds and patient capital at infrastructure scale</li><li>Fast-track permitting and abundant cheap energy</li><li>Vision 2030 and national AI strategies as industrial policy</li><li>EMEA projected to add 13 GW by 2030 — Gulf taking share from Europe</li></ul><p><strong>Related keywords:</strong> Middle East data centers, GCC AI investment, Saudi Arabia Humain, Stargate UAE, Abu Dhabi data center, sovereign wealth funds, Vision 2030, international AI infrastructure</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>AI data center investments across the Gulf Cooperation Council countries are projected to exceed $5 billion in 2026. The Middle East has entered the AI infrastructure race with structural advantages that constrained Western markets cannot match.</p><p><strong>Key topics covered:</strong></p><ul><li>Saudi Arabia's Humain: $1.2 billion for 250 megawatts of AI capacity</li><li>OpenAI Stargate UAE: 200MW first phase of a 1 gigawatt cluster</li><li>Sovereign wealth funds and patient capital at infrastructure scale</li><li>Fast-track permitting and abundant cheap energy</li><li>Vision 2030 and national AI strategies as industrial policy</li><li>EMEA projected to add 13 GW by 2030 — Gulf taking share from Europe</li></ul><p><strong>Related keywords:</strong> Middle East data centers, GCC AI investment, Saudi Arabia Humain, Stargate UAE, Abu Dhabi data center, sovereign wealth funds, Vision 2030, international AI infrastructure</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </content:encoded>
      <pubDate>Wed, 28 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/96cc2e0a/54c02b36.mp3" length="2635879" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>322</itunes:duration>
      <itunes:summary>Saudi Arabia's Humain secures $1.2B for 250MW. OpenAI's Stargate brings 200MW online in Abu Dhabi. The Gulf is moving fast on AI infrastructure.</itunes:summary>
      <itunes:subtitle>Saudi Arabia's Humain secures $1.2B for 250MW. OpenAI's Stargate brings 200MW online in Abu Dhabi. The Gulf is moving fast on AI infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>OpenAI Says the Grid Can't Keep Up</title>
      <itunes:episode>20</itunes:episode>
      <podcast:episode>20</podcast:episode>
      <itunes:title>OpenAI Says the Grid Can't Keep Up</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">fbc696b1-39d6-4e44-8127-a3ba3caae9ff</guid>
      <link>https://share.transistor.fm/s/91edc8ff</link>
      <description>
        <![CDATA[<p>OpenAI is publicly warning that the U.S. electrical grid cannot support the pace of AI development. The company estimates the nation needs approximately 100 gigawatts of new generating capacity annually — a buildout rate not seen since post-war electrification.</p><p><strong>Key topics covered:</strong></p><ul><li>Data center power consumption projected to reach 123 GW by 2035 (up from 4 GW in 2024)</li><li>Infrastructure bottlenecks: transformers, interconnection queues, transmission permitting</li><li>Behind-the-meter solutions and on-site generation strategies</li><li>Stargate, xAI Memphis, and Three Mile Island restart</li><li>Grid-dependent vs. grid-independent development risk</li><li>The premium for shovel-ready sites with secured power</li></ul><p><strong>Related keywords:</strong> grid capacity, AI power demand, behind-the-meter generation, interconnection queue, data center power, OpenAI infrastructure, Stargate project, utility constraints</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>OpenAI is publicly warning that the U.S. electrical grid cannot support the pace of AI development. The company estimates the nation needs approximately 100 gigawatts of new generating capacity annually — a buildout rate not seen since post-war electrification.</p><p><strong>Key topics covered:</strong></p><ul><li>Data center power consumption projected to reach 123 GW by 2035 (up from 4 GW in 2024)</li><li>Infrastructure bottlenecks: transformers, interconnection queues, transmission permitting</li><li>Behind-the-meter solutions and on-site generation strategies</li><li>Stargate, xAI Memphis, and Three Mile Island restart</li><li>Grid-dependent vs. grid-independent development risk</li><li>The premium for shovel-ready sites with secured power</li></ul><p><strong>Related keywords:</strong> grid capacity, AI power demand, behind-the-meter generation, interconnection queue, data center power, OpenAI infrastructure, Stargate project, utility constraints</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </content:encoded>
      <pubDate>Tue, 27 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/91edc8ff/c3164624.mp3" length="1684171" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>203</itunes:duration>
      <itunes:summary>OpenAI warns the U.S. needs 100 gigawatts of new capacity annually. Data center power consumption could hit 123 GW by 2035 — a 30x increase.</itunes:summary>
      <itunes:subtitle>OpenAI warns the U.S. needs 100 gigawatts of new capacity annually. Data center power consumption could hit 123 GW by 2035 — a 30x increase.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Water Problem No One Wants to Talk About</title>
      <itunes:episode>19</itunes:episode>
      <podcast:episode>19</podcast:episode>
      <itunes:title>The Water Problem No One Wants to Talk About</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">db5aac54-a5a9-4a04-a5bb-ed8432439d37</guid>
      <link>https://share.transistor.fm/s/32330076</link>
      <description>
        <![CDATA[<p>A medium-sized data center consumes approximately 110 million gallons of water per year — equivalent to the annual usage of 1,000 households. As AI infrastructure scales, water is emerging as the second binding constraint after electricity, generating community opposition that developers can no longer ignore.</p><p><strong>Key topics covered:</strong></p><ul><li>U.S. data center water consumption projected to surge to 68 billion gallons annually by 2028</li><li>Evaporative cooling mechanics and why water doesn't recirculate</li><li>Site selection tension: energy availability vs. water availability</li><li>Community pushback and state-level regulatory responses</li><li>Closed-loop cooling, liquid cooling, and water restoration programs</li><li>Microsoft, Google, and Meta water-positive commitments</li></ul><p><strong>Related keywords:</strong> data center water usage, AI water consumption, evaporative cooling, water-stressed regions, closed-loop cooling, immersion cooling, water positive data centers, sustainability</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>A medium-sized data center consumes approximately 110 million gallons of water per year — equivalent to the annual usage of 1,000 households. As AI infrastructure scales, water is emerging as the second binding constraint after electricity, generating community opposition that developers can no longer ignore.</p><p><strong>Key topics covered:</strong></p><ul><li>U.S. data center water consumption projected to surge to 68 billion gallons annually by 2028</li><li>Evaporative cooling mechanics and why water doesn't recirculate</li><li>Site selection tension: energy availability vs. water availability</li><li>Community pushback and state-level regulatory responses</li><li>Closed-loop cooling, liquid cooling, and water restoration programs</li><li>Microsoft, Google, and Meta water-positive commitments</li></ul><p><strong>Related keywords:</strong> data center water usage, AI water consumption, evaporative cooling, water-stressed regions, closed-loop cooling, immersion cooling, water positive data centers, sustainability</p><p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]>
      </content:encoded>
      <pubDate>Mon, 26 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/32330076/194be6de.mp3" length="2730960" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>334</itunes:duration>
      <itunes:summary>A medium-sized data center consumes 110 million gallons of water per year. As AI scales, water is emerging as the second binding constraint after electricity.</itunes:summary>
      <itunes:subtitle>A medium-sized data center consumes 110 million gallons of water per year. As AI scales, water is emerging as the second binding constraint after electricity.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>xAI's $20 Billion Bet on Mississippi</title>
      <itunes:episode>18</itunes:episode>
      <podcast:episode>18</podcast:episode>
      <itunes:title>xAI's $20 Billion Bet on Mississippi</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">e52fb596-0142-4ce5-a674-142d3fd8d81b</guid>
      <link>https://share.transistor.fm/s/4f36f400</link>
      <description>
        <![CDATA[Elon Musk's xAI just announced the largest private investment in Mississippi history: a $20 billion AI supercomputer facility with 2 gigawatts of power capacity. In this episode of The Power Allocation, we analyze why megascale AI infrastructure is moving to where the electrons are.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>xAI's $20 billion Mississippi AI training facility</li>
<li>Why 2 gigawatts equals the output of two nuclear reactors</li>
<li>The Memphis 1 gigawatt cluster — the first of its kind globally</li>
<li>Why Mississippi beat Northern Virginia and Silicon Valley for this project</li>
<li>xAI's pursuit of direct nuclear power agreements</li>
<li>How AI infrastructure capital formation now rivals heavy industry</li>
</ul>

<p><strong>Related keywords:</strong> xAI Mississippi data center, Elon Musk AI infrastructure, xAI $20 billion investment, Southaven Mississippi data center, xAI Memphis colossus, 2 gigawatt data center, AI supercomputer facility, xAI Series E funding, AI factory power consumption, megascale AI infrastructure, data center site selection</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Elon Musk's xAI just announced the largest private investment in Mississippi history: a $20 billion AI supercomputer facility with 2 gigawatts of power capacity. In this episode of The Power Allocation, we analyze why megascale AI infrastructure is moving to where the electrons are.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>xAI's $20 billion Mississippi AI training facility</li>
<li>Why 2 gigawatts equals the output of two nuclear reactors</li>
<li>The Memphis 1 gigawatt cluster — the first of its kind globally</li>
<li>Why Mississippi beat Northern Virginia and Silicon Valley for this project</li>
<li>xAI's pursuit of direct nuclear power agreements</li>
<li>How AI infrastructure capital formation now rivals heavy industry</li>
</ul>

<p><strong>Related keywords:</strong> xAI Mississippi data center, Elon Musk AI infrastructure, xAI $20 billion investment, Southaven Mississippi data center, xAI Memphis colossus, 2 gigawatt data center, AI supercomputer facility, xAI Series E funding, AI factory power consumption, megascale AI infrastructure, data center site selection</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Fri, 23 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/4f36f400/8807c12d.mp3" length="2253642" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>275</itunes:duration>
      <itunes:summary>xAI announced a $20 billion AI supercomputer facility in Mississippi — the largest private investment in state history. Initial capacity: 2 gigawatts. Why megascale AI infrastructure is going where the power is.</itunes:summary>
      <itunes:subtitle>xAI announced a $20 billion AI supercomputer facility in Mississippi — the largest private investment in state history. Initial capacity: 2 gigawatts. Why megascale AI infrastructure is going where the power is.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Liquid Cooling Becomes Mandatory</title>
      <itunes:episode>17</itunes:episode>
      <podcast:episode>17</podcast:episode>
      <itunes:title>Liquid Cooling Becomes Mandatory</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">a51aa63b-9b80-4e00-8942-01ef74e06d07</guid>
      <link>https://share.transistor.fm/s/292aa244</link>
      <description>
        <![CDATA[Air cooling can no longer keep up with AI. The data center liquid cooling market is growing at 50% annually and has already captured 46% of the market. In this episode of The Power Allocation, we explain why direct-to-chip and immersion cooling are now mandatory for AI infrastructure.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why AI accelerators generate heat densities air cooling cannot handle</li>
<li>The $11 billion liquid cooling market projected by 2030</li>
<li>Direct-to-chip vs. immersion cooling technologies</li>
<li>NVIDIA's hot water cooling architecture at 45°C</li>
<li>Why legacy data centers cannot compete for hyperscaler AI workloads</li>
<li>The difference between "AI-ready" marketing and actual AI-capable facilities</li>
</ul>

<p><strong>Related keywords:</strong> data center liquid cooling, AI chip cooling, direct to chip cooling, immersion cooling data center, NVIDIA liquid cooling, data center thermal management, AI accelerator heat density, data center cooling market, rear door heat exchanger, dielectric fluid cooling, AI-ready data center</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Air cooling can no longer keep up with AI. The data center liquid cooling market is growing at 50% annually and has already captured 46% of the market. In this episode of The Power Allocation, we explain why direct-to-chip and immersion cooling are now mandatory for AI infrastructure.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why AI accelerators generate heat densities air cooling cannot handle</li>
<li>The $11 billion liquid cooling market projected by 2030</li>
<li>Direct-to-chip vs. immersion cooling technologies</li>
<li>NVIDIA's hot water cooling architecture at 45°C</li>
<li>Why legacy data centers cannot compete for hyperscaler AI workloads</li>
<li>The difference between "AI-ready" marketing and actual AI-capable facilities</li>
</ul>

<p><strong>Related keywords:</strong> data center liquid cooling, AI chip cooling, direct to chip cooling, immersion cooling data center, NVIDIA liquid cooling, data center thermal management, AI accelerator heat density, data center cooling market, rear door heat exchanger, dielectric fluid cooling, AI-ready data center</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Thu, 22 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/292aa244/9961e9bc.mp3" length="2447571" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>299</itunes:duration>
      <itunes:summary>Liquid cooling has captured 46% of the data center market and is growing at 50% annually. Why direct-to-chip and immersion cooling are now mandatory for AI infrastructure.</itunes:summary>
      <itunes:subtitle>Liquid cooling has captured 46% of the data center market and is growing at 50% annually. Why direct-to-chip and immersion cooling are now mandatory for AI infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Political Backlash Against AI Power Demand Is Here</title>
      <itunes:episode>16</itunes:episode>
      <podcast:episode>16</podcast:episode>
      <itunes:title>The Political Backlash Against AI Power Demand Is Here</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">bef44bb1-11c8-41fa-938f-d754304c348e</guid>
      <link>https://share.transistor.fm/s/3ad96740</link>
      <description>
        <![CDATA[The political backlash against AI's power consumption has arrived. Bernie Sanders wants a moratorium on data centers. Ron DeSantis is blocking AI expansion in Florida. And Tom Cotton's DATA Act could let facilities build their own off-grid power. In this episode of The Power Allocation, we map the emerging political fault lines around AI infrastructure.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Senator Sanders' call for a data center construction moratorium</li>
<li>Governor DeSantis' resistance to AI facility expansion in Florida</li>
<li>The DATA Act and its implications for behind-the-meter generation</li>
<li>Why AI data centers could consume 5% of U.S. electricity by 2028</li>
<li>How regulatory risk is reshaping data center site selection</li>
</ul>

<p><strong>Related keywords:</strong> data center regulation, AI power consumption politics, Bernie Sanders data center moratorium, DeSantis AI expansion, DATA Act Tom Cotton, data center electricity demand, grid strain AI, data center site selection risk, behind the meter generation, data center political opposition</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[The political backlash against AI's power consumption has arrived. Bernie Sanders wants a moratorium on data centers. Ron DeSantis is blocking AI expansion in Florida. And Tom Cotton's DATA Act could let facilities build their own off-grid power. In this episode of The Power Allocation, we map the emerging political fault lines around AI infrastructure.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Senator Sanders' call for a data center construction moratorium</li>
<li>Governor DeSantis' resistance to AI facility expansion in Florida</li>
<li>The DATA Act and its implications for behind-the-meter generation</li>
<li>Why AI data centers could consume 5% of U.S. electricity by 2028</li>
<li>How regulatory risk is reshaping data center site selection</li>
</ul>

<p><strong>Related keywords:</strong> data center regulation, AI power consumption politics, Bernie Sanders data center moratorium, DeSantis AI expansion, DATA Act Tom Cotton, data center electricity demand, grid strain AI, data center site selection risk, behind the meter generation, data center political opposition</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Wed, 21 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/3ad96740/ebc3b023.mp3" length="2127437" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>259</itunes:duration>
      <itunes:summary>Bernie Sanders calls for a moratorium on data centers. Ron DeSantis blocks AI expansion. Tom Cotton proposes letting facilities go off-grid. The political battle over AI's power demand has begun.</itunes:summary>
      <itunes:subtitle>Bernie Sanders calls for a moratorium on data centers. Ron DeSantis blocks AI expansion. Tom Cotton proposes letting facilities go off-grid. The political battle over AI's power demand has begun.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why Transformers Are the New GPUs</title>
      <itunes:episode>15</itunes:episode>
      <podcast:episode>15</podcast:episode>
      <itunes:title>Why Transformers Are the New GPUs</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">ee2f9173-b98f-4218-aa08-8e1861265bb4</guid>
      <link>https://share.transistor.fm/s/0c5ddc34</link>
      <description>
        <![CDATA[Power transformers have become the unexpected bottleneck in AI infrastructure. Lead times now exceed two years, and data centers across Silicon Valley sit dark — fully built but unable to operate. In this episode of The Power Allocation, we examine why electrical equipment is the new constraint on AI growth.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why transformer lead times have stretched to 24+ months</li>
<li>The 70% price increase in transformer costs since 2019</li>
<li>How the U.S. imports 80% of large high-voltage transformers</li>
<li>Why half of America's distribution transformers are past service life</li>
<li>Capital implications for data center developers and infrastructure investors</li>
</ul>

<p><strong>Related keywords:</strong> power transformer shortage, data center electrical infrastructure, transformer lead times, high voltage transformer supply chain, electrical steel shortage, data center grid connection, AI infrastructure bottleneck, transformer manufacturing capacity, data center power delivery, electrical equipment backlog</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Power transformers have become the unexpected bottleneck in AI infrastructure. Lead times now exceed two years, and data centers across Silicon Valley sit dark — fully built but unable to operate. In this episode of The Power Allocation, we examine why electrical equipment is the new constraint on AI growth.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why transformer lead times have stretched to 24+ months</li>
<li>The 70% price increase in transformer costs since 2019</li>
<li>How the U.S. imports 80% of large high-voltage transformers</li>
<li>Why half of America's distribution transformers are past service life</li>
<li>Capital implications for data center developers and infrastructure investors</li>
</ul>

<p><strong>Related keywords:</strong> power transformer shortage, data center electrical infrastructure, transformer lead times, high voltage transformer supply chain, electrical steel shortage, data center grid connection, AI infrastructure bottleneck, transformer manufacturing capacity, data center power delivery, electrical equipment backlog</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Tue, 20 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/0c5ddc34/60ceec7c.mp3" length="2322394" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>283</itunes:duration>
      <itunes:summary>Lead times for power transformers now exceed two years. Two Silicon Valley data centers sit dark, fully built but waiting on electrical equipment. Why the transformer shortage is the next bottleneck.</itunes:summary>
      <itunes:subtitle>Lead times for power transformers now exceed two years. Two Silicon Valley data centers sit dark, fully built but waiting on electrical equipment. Why the transformer shortage is the next bottleneck.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Meta Becomes the Biggest Nuclear Buyer in Big Tech</title>
      <itunes:episode>14</itunes:episode>
      <podcast:episode>14</podcast:episode>
      <itunes:title>Meta Becomes the Biggest Nuclear Buyer in Big Tech</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">1d318de7-d4ba-4cad-9ed7-92c4b626b690</guid>
      <link>https://share.transistor.fm/s/d38f533d</link>
      <description>
        <![CDATA[Meta just became the largest corporate buyer of nuclear power in tech history. In this episode of The Power Allocation, we break down Meta's 6.6 gigawatt nuclear energy deal with Vistra Corp, TerraPower, and Oklo — and what it means for AI infrastructure investment.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Meta chose nuclear over renewables for AI data center power</li>
<li>How nuclear's 90%+ capacity factor solves AI's baseload problem</li>
<li>The role of small modular reactors (SMRs) in next-generation data centers</li>
<li>Investment implications for Vistra, Constellation Energy, Oklo, NuScale, and TerraPower</li>
<li>Why hyperscalers are racing to lock up America's nuclear capacity</li>
</ul>

<p><strong>Related keywords:</strong> Meta nuclear power deal, AI data center energy, Vistra Corp AI, TerraPower data center, Oklo SMR, nuclear power AI infrastructure, hyperscaler energy procurement, baseload power AI, data center power purchase agreement, Microsoft Constellation, Amazon SMR investment</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Meta just became the largest corporate buyer of nuclear power in tech history. In this episode of The Power Allocation, we break down Meta's 6.6 gigawatt nuclear energy deal with Vistra Corp, TerraPower, and Oklo — and what it means for AI infrastructure investment.

<p><strong>Key topics covered:</strong></p>
<ul>
<li>Why Meta chose nuclear over renewables for AI data center power</li>
<li>How nuclear's 90%+ capacity factor solves AI's baseload problem</li>
<li>The role of small modular reactors (SMRs) in next-generation data centers</li>
<li>Investment implications for Vistra, Constellation Energy, Oklo, NuScale, and TerraPower</li>
<li>Why hyperscalers are racing to lock up America's nuclear capacity</li>
</ul>

<p><strong>Related keywords:</strong> Meta nuclear power deal, AI data center energy, Vistra Corp AI, TerraPower data center, Oklo SMR, nuclear power AI infrastructure, hyperscaler energy procurement, baseload power AI, data center power purchase agreement, Microsoft Constellation, Amazon SMR investment</p>

<p>The Power Allocation is brought to you by Spring Street Management Group — translating AI and data center hype into real infrastructure and assets.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Mon, 19 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/d38f533d/d6137970.mp3" length="2299841" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>280</itunes:duration>
      <itunes:summary>Meta signed the largest nuclear power deal in tech history — 6.6 gigawatts from Vistra, TerraPower, and Oklo. Why nuclear is becoming the strategic energy play for AI infrastructure.</itunes:summary>
      <itunes:subtitle>Meta signed the largest nuclear power deal in tech history — 6.6 gigawatts from Vistra, TerraPower, and Oklo. Why nuclear is becoming the strategic energy play for AI infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Case for Grid-Independent AI Infrastructure</title>
      <itunes:episode>13</itunes:episode>
      <podcast:episode>13</podcast:episode>
      <itunes:title>The Case for Grid-Independent AI Infrastructure</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">47556c8e-718e-499a-98e0-2c53c5a2dc10</guid>
      <link>https://share.transistor.fm/s/540dc5ef</link>
      <description>
        <![CDATA[What if the solution to grid constraints is to skip the grid entirely?<p>GridFree AI, a Houston-based startup, announced its first grid-independent site in South Dallas — part of a planned three-site campus with nearly 5 gigawatts of combined capacity. Five gigawatts. That's more than many countries consume. And none of it connected to the public grid.</p><p>The model: on-site power generation using natural gas, with potential pathways to hydrogen and other fuels. No interconnection queue. No utility negotiations. No 4-7 year wait times.</p><p>Grid-independent infrastructure solves the binding constraint of the AI buildout — time. The tradeoff is cost and complexity, but for operators who need capacity now, the premium may be worth paying.</p><p>The grid was built for a different era. Some developers aren't waiting for it to catch up.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[What if the solution to grid constraints is to skip the grid entirely?<p>GridFree AI, a Houston-based startup, announced its first grid-independent site in South Dallas — part of a planned three-site campus with nearly 5 gigawatts of combined capacity. Five gigawatts. That's more than many countries consume. And none of it connected to the public grid.</p><p>The model: on-site power generation using natural gas, with potential pathways to hydrogen and other fuels. No interconnection queue. No utility negotiations. No 4-7 year wait times.</p><p>Grid-independent infrastructure solves the binding constraint of the AI buildout — time. The tradeoff is cost and complexity, but for operators who need capacity now, the premium may be worth paying.</p><p>The grid was built for a different era. Some developers aren't waiting for it to catch up.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Fri, 16 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/540dc5ef/35c6cb1b.mp3" length="1789719" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>217</itunes:duration>
      <itunes:summary>What if the solution to grid constraints is to skip the grid entirely? GridFree AI's 5-gigawatt model and the rise of on-site generation for AI compute.</itunes:summary>
      <itunes:subtitle>What if the solution to grid constraints is to skip the grid entirely? GridFree AI's 5-gigawatt model and the rise of on-site generation for AI compute.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Inside Stargate — The 902-Megawatt Bet on AI Infrastructure</title>
      <itunes:episode>12</itunes:episode>
      <podcast:episode>12</podcast:episode>
      <itunes:title>Inside Stargate — The 902-Megawatt Bet on AI Infrastructure</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">d3e698dd-6fde-4348-bf2e-ebbbdf9b73af</guid>
      <link>https://share.transistor.fm/s/5d7a0aae</link>
      <description>
        <![CDATA[Oracle and OpenAI aren't just building a data center. They're building a power plant that happens to run GPUs.<p>Vantage Data Centers broke ground on the Lighthouse project in Port Washington, Wisconsin — a $15 billion campus comprising four data centers delivering 902 megawatts of IT capacity. This is part of the Stargate initiative, the joint venture between Oracle, OpenAI, and SoftBank. To put 902 megawatts in perspective: that's roughly equivalent to the output of a small nuclear reactor. It's enough to power 700,000 homes.</p><p>Wisconsin wasn't chosen for its tech ecosystem. It was chosen for power. The state has available grid capacity, favorable utility relationships, and transmission infrastructure that can actually deliver at this scale.</p><p>$15 billion for a single campus signals the capital intensity of the AI buildout. This isn't venture scale — it's infrastructure scale. The AI race is now a construction race.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Oracle and OpenAI aren't just building a data center. They're building a power plant that happens to run GPUs.<p>Vantage Data Centers broke ground on the Lighthouse project in Port Washington, Wisconsin — a $15 billion campus comprising four data centers delivering 902 megawatts of IT capacity. This is part of the Stargate initiative, the joint venture between Oracle, OpenAI, and SoftBank. To put 902 megawatts in perspective: that's roughly equivalent to the output of a small nuclear reactor. It's enough to power 700,000 homes.</p><p>Wisconsin wasn't chosen for its tech ecosystem. It was chosen for power. The state has available grid capacity, favorable utility relationships, and transmission infrastructure that can actually deliver at this scale.</p><p>$15 billion for a single campus signals the capital intensity of the AI buildout. This isn't venture scale — it's infrastructure scale. The AI race is now a construction race.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Thu, 15 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/5d7a0aae/aba8db53.mp3" length="1819676" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>220</itunes:duration>
      <itunes:summary>Oracle and OpenAI aren't just building a data center. They're building a power plant that happens to run GPUs. Inside the $15 billion Lighthouse project in Wisconsin.</itunes:summary>
      <itunes:subtitle>Oracle and OpenAI aren't just building a data center. They're building a power plant that happens to run GPUs. Inside the $15 billion Lighthouse project in Wisconsin.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why Google Just Paid $4.75 Billion for a Power Company</title>
      <itunes:episode>11</itunes:episode>
      <podcast:episode>11</podcast:episode>
      <itunes:title>Why Google Just Paid $4.75 Billion for a Power Company</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">d6b330f7-fcc1-4a2b-a850-5fc6a1e6d392</guid>
      <link>https://share.transistor.fm/s/204b2de0</link>
      <description>
        <![CDATA[Google didn't just buy a clean energy company. It bought a shortcut to power.<p>Alphabet acquired Intersect Power for $4.75 billion in cash, assuming existing debt. Intersect isn't a tech company — it's a developer of utility-scale solar, storage, and data center-ready power infrastructure. Google already had one of the largest corporate renewable portfolios in the world. This acquisition isn't about going green. It's about going faster.</p><p>Grid interconnection queues now stretch 4-7 years in major markets. Intersect already has operational assets, development pipeline, and — critically — interconnection agreements already in place. Google isn't buying megawatts. It's buying queue position. It's buying time.</p><p>This deal signals a phase shift in how hyperscalers think about infrastructure. Vertical integration into power generation is no longer optional — it's strategic. Power ownership is becoming a moat.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[Google didn't just buy a clean energy company. It bought a shortcut to power.<p>Alphabet acquired Intersect Power for $4.75 billion in cash, assuming existing debt. Intersect isn't a tech company — it's a developer of utility-scale solar, storage, and data center-ready power infrastructure. Google already had one of the largest corporate renewable portfolios in the world. This acquisition isn't about going green. It's about going faster.</p><p>Grid interconnection queues now stretch 4-7 years in major markets. Intersect already has operational assets, development pipeline, and — critically — interconnection agreements already in place. Google isn't buying megawatts. It's buying queue position. It's buying time.</p><p>This deal signals a phase shift in how hyperscalers think about infrastructure. Vertical integration into power generation is no longer optional — it's strategic. Power ownership is becoming a moat.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Wed, 14 Jan 2026 06:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/204b2de0/202ca248.mp3" length="2154395" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>262</itunes:duration>
      <itunes:summary>Google didn't just buy a clean energy company. It bought a shortcut to power. Alphabet's $4.75 billion acquisition of Intersect Power signals vertical integration into generation is now strategic.</itunes:summary>
      <itunes:subtitle>Google didn't just buy a clean energy company. It bought a shortcut to power. Alphabet's $4.75 billion acquisition of Intersect Power signals vertical integration into generation is now strategic.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Time Lag Capital Underestimates in Power Delivery</title>
      <itunes:episode>10</itunes:episode>
      <podcast:episode>10</podcast:episode>
      <itunes:title>The Time Lag Capital Underestimates in Power Delivery</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">91a89a2d-2553-4a50-8701-4f3a00060ae1</guid>
      <link>https://share.transistor.fm/s/16683c08</link>
      <description>
        <![CDATA[<p>Capital can be deployed in months. Power infrastructure takes years. This mismatch is the defining tension of the AI buildout.</p><p>A 200-megawatt data center requires 12-24 months for land and permitting, 18-24 months for construction, and 24-48 months for grid interconnection. Money raised in 2024 may not translate into running compute until 2028 or later.</p><p>The AI infrastructure winners will be determined by decisions made in 2020-2023, not decisions being made now. Capital moves fast. Infrastructure moves on its own schedule.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Capital can be deployed in months. Power infrastructure takes years. This mismatch is the defining tension of the AI buildout.</p><p>A 200-megawatt data center requires 12-24 months for land and permitting, 18-24 months for construction, and 24-48 months for grid interconnection. Money raised in 2024 may not translate into running compute until 2028 or later.</p><p>The AI infrastructure winners will be determined by decisions made in 2020-2023, not decisions being made now. Capital moves fast. Infrastructure moves on its own schedule.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Tue, 13 Jan 2026 07:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/16683c08/2304072b.mp3" length="1421294" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>171</itunes:duration>
      <itunes:summary>Capital can be deployed in months. Power infrastructure takes years. This mismatch is the defining tension of the AI buildout.</itunes:summary>
      <itunes:subtitle>Capital can be deployed in months. Power infrastructure takes years. This mismatch is the defining tension of the AI buildout.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why Renewable Power Alone Isn't Enough</title>
      <itunes:episode>9</itunes:episode>
      <podcast:episode>9</podcast:episode>
      <itunes:title>Why Renewable Power Alone Isn't Enough</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">148cc186-b0eb-4ca5-9112-49451b3f351b</guid>
      <link>https://share.transistor.fm/s/428c4696</link>
      <description>
        <![CDATA[<p>Every hyperscaler has announced ambitious renewable energy targets. But renewables alone cannot power the AI buildout. Here's why.</p><p>AI workloads run continuously. Solar generates power roughly 25% of the time. Wind achieves 35-45% capacity factors. Neither provides the baseload reliability that AI facilities require. Battery storage helps but remains expensive at scale.</p><p>The AI buildout is driving investment in both renewables and dispatchable generation. Clean energy is a goal. Reliable power is a requirement.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Every hyperscaler has announced ambitious renewable energy targets. But renewables alone cannot power the AI buildout. Here's why.</p><p>AI workloads run continuously. Solar generates power roughly 25% of the time. Wind achieves 35-45% capacity factors. Neither provides the baseload reliability that AI facilities require. Battery storage helps but remains expensive at scale.</p><p>The AI buildout is driving investment in both renewables and dispatchable generation. Clean energy is a goal. Reliable power is a requirement.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Mon, 12 Jan 2026 07:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/428c4696/69185617.mp3" length="1131005" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>134</itunes:duration>
      <itunes:summary>Every hyperscaler has announced ambitious renewable energy targets. But renewables alone cannot power the AI buildout. Here's why.</itunes:summary>
      <itunes:subtitle>Every hyperscaler has announced ambitious renewable energy targets. But renewables alone cannot power the AI buildout. Here's why.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why Utilities Are the Quiet Kingmakers of AI</title>
      <itunes:episode>8</itunes:episode>
      <podcast:episode>8</podcast:episode>
      <itunes:title>Why Utilities Are the Quiet Kingmakers of AI</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">ffedd9f2-3c8c-486e-8894-fe300eb29e00</guid>
      <link>https://share.transistor.fm/s/1d0aa6be</link>
      <description>
        <![CDATA[<p>The AI buildout depends on decisions made by regulated utilities — entities most tech investors have never studied. Utilities are the quiet kingmakers of this cycle.</p><p>Utilities operate under regulatory frameworks that prioritize reliability and rate stability. Adding 200 megawatts of load for a single customer requires new substations, transmission upgrades, and generation procurement — all approved by state regulators.</p><p>Utility strategy is becoming a differentiator in AI infrastructure deployment. Companies that understand how to work within regulated frameworks have structural advantages that pure capital cannot replicate.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>The AI buildout depends on decisions made by regulated utilities — entities most tech investors have never studied. Utilities are the quiet kingmakers of this cycle.</p><p>Utilities operate under regulatory frameworks that prioritize reliability and rate stability. Adding 200 megawatts of load for a single customer requires new substations, transmission upgrades, and generation procurement — all approved by state regulators.</p><p>Utility strategy is becoming a differentiator in AI infrastructure deployment. Companies that understand how to work within regulated frameworks have structural advantages that pure capital cannot replicate.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Fri, 09 Jan 2026 07:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/1d0aa6be/c32c267b.mp3" length="1478335" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>178</itunes:duration>
      <itunes:summary>The AI buildout depends on decisions made by regulated utilities — entities most tech investors have never studied. Utilities are the quiet kingmakers of this cycle.</itunes:summary>
      <itunes:subtitle>The AI buildout depends on decisions made by regulated utilities — entities most tech investors have never studied. Utilities are the quiet kingmakers of this cycle.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>How Power Pricing Shapes Data Center Geography</title>
      <itunes:episode>7</itunes:episode>
      <podcast:episode>7</podcast:episode>
      <itunes:title>How Power Pricing Shapes Data Center Geography</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">6953e3c7-1a65-426a-a592-7970025f7a3b</guid>
      <link>https://share.transistor.fm/s/cc625f93</link>
      <description>
        <![CDATA[<p>Data centers don't just need power — they need cheap power. And the geography of electricity pricing is reshaping where AI infrastructure gets built.</p><p>For a 100-megawatt facility running 24/7, each cent per kilowatt-hour represents roughly $8.7 million in annual operating cost. A 5-cent differential means $43 million per year — nearly a billion dollars over a 20-year horizon.</p><p>Power pricing is creating durable geographic advantages that no amount of tax credits can offset. Capital follows the electrons.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Data centers don't just need power — they need cheap power. And the geography of electricity pricing is reshaping where AI infrastructure gets built.</p><p>For a 100-megawatt facility running 24/7, each cent per kilowatt-hour represents roughly $8.7 million in annual operating cost. A 5-cent differential means $43 million per year — nearly a billion dollars over a 20-year horizon.</p><p>Power pricing is creating durable geographic advantages that no amount of tax credits can offset. Capital follows the electrons.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Thu, 08 Jan 2026 07:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/cc625f93/57e98292.mp3" length="1510311" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>182</itunes:duration>
      <itunes:summary>Data centers don't just need power — they need cheap power. And the geography of electricity pricing is reshaping where AI infrastructure gets built.</itunes:summary>
      <itunes:subtitle>Data centers don't just need power — they need cheap power. And the geography of electricity pricing is reshaping where AI infrastructure gets built.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why Grid Interconnection Is the Real Bottleneck</title>
      <itunes:episode>6</itunes:episode>
      <podcast:episode>6</podcast:episode>
      <itunes:title>Why Grid Interconnection Is the Real Bottleneck</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">e28482d0-2221-4691-a59d-c4201a945c10</guid>
      <link>https://share.transistor.fm/s/3433ec7d</link>
      <description>
        <![CDATA[<p>You can buy land. You can order GPUs. You can raise capital. But you cannot buy a faster grid connection. Interconnection is the binding constraint on AI infrastructure.</p><p>In PJM — the grid operator for the Mid-Atlantic — the interconnection queue now exceeds 2,500 projects totaling over 250 gigawatts. The average wait time has stretched beyond 4 years. Developers who secured interconnection agreements 3-5 years ago now hold strategic assets.</p><p>Capital is repricing assets based on interconnection status, not just location or capacity. The race for AI compute is increasingly a race for queue position.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>You can buy land. You can order GPUs. You can raise capital. But you cannot buy a faster grid connection. Interconnection is the binding constraint on AI infrastructure.</p><p>In PJM — the grid operator for the Mid-Atlantic — the interconnection queue now exceeds 2,500 projects totaling over 250 gigawatts. The average wait time has stretched beyond 4 years. Developers who secured interconnection agreements 3-5 years ago now hold strategic assets.</p><p>Capital is repricing assets based on interconnection status, not just location or capacity. The race for AI compute is increasingly a race for queue position.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Wed, 07 Jan 2026 07:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/3433ec7d/21a436d0.mp3" length="1287749" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>154</itunes:duration>
      <itunes:summary>You can buy land. You can order GPUs. You can raise capital. But you cannot buy a faster grid connection. Interconnection is the binding constraint on AI infrastructure.</itunes:summary>
      <itunes:subtitle>You can buy land. You can order GPUs. You can raise capital. But you cannot buy a faster grid connection. Interconnection is the binding constraint on AI infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>What Hyperscalers Care About That Startups Don't</title>
      <itunes:episode>5</itunes:episode>
      <podcast:episode>5</podcast:episode>
      <itunes:title>What Hyperscalers Care About That Startups Don't</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">a8489008-54db-47a1-9bd6-504825ae72f7</guid>
      <link>https://share.transistor.fm/s/00e896ff</link>
      <description>
        <![CDATA[<p>Hyperscalers and AI startups operate in the same industry but play entirely different games. The difference is infrastructure.</p><p>Hyperscalers sign 10-15 year power purchase agreements, coordinate grid interconnection at utility scale, deploy liquid cooling systems, and plan 5-10 years ahead. Microsoft has nuclear agreements. Google invests in geothermal. Amazon backs utility-scale renewables.</p><p>Startups rent the result. Hyperscalers own the moat. In AI infrastructure, scale isn't an advantage — it's the only advantage.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Hyperscalers and AI startups operate in the same industry but play entirely different games. The difference is infrastructure.</p><p>Hyperscalers sign 10-15 year power purchase agreements, coordinate grid interconnection at utility scale, deploy liquid cooling systems, and plan 5-10 years ahead. Microsoft has nuclear agreements. Google invests in geothermal. Amazon backs utility-scale renewables.</p><p>Startups rent the result. Hyperscalers own the moat. In AI infrastructure, scale isn't an advantage — it's the only advantage.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Tue, 06 Jan 2026 07:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/00e896ff/3d52b752.mp3" length="1175482" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>147</itunes:duration>
      <itunes:summary>Hyperscalers and AI startups operate in the same industry but play entirely different games. The difference is infrastructure.</itunes:summary>
      <itunes:subtitle>Hyperscalers and AI startups operate in the same industry but play entirely different games. The difference is infrastructure.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why Land Is Back in the AI Equation</title>
      <itunes:episode>4</itunes:episode>
      <podcast:episode>4</podcast:episode>
      <itunes:title>Why Land Is Back in the AI Equation</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">582cd185-e1be-4c95-b390-cde5de75e04f</guid>
      <link>https://share.transistor.fm/s/1aad0c3a</link>
      <description>
        <![CDATA[<p>For a decade, software ate the world. Now AI is making land matter again.</p><p>Meta's 2-gigawatt facility requires hundreds of acres. Microsoft's expansion spans millions of square feet globally. These aren't software deployments — they're industrial developments with 30-year time horizons requiring transmission adjacency, water rights, and favorable zoning.</p><p>Developers with optioned land near transmission corridors are commanding premium valuations. Land isn't just back — it's becoming a rate-limiting input to the AI buildout.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>For a decade, software ate the world. Now AI is making land matter again.</p><p>Meta's 2-gigawatt facility requires hundreds of acres. Microsoft's expansion spans millions of square feet globally. These aren't software deployments — they're industrial developments with 30-year time horizons requiring transmission adjacency, water rights, and favorable zoning.</p><p>Developers with optioned land near transmission corridors are commanding premium valuations. Land isn't just back — it's becoming a rate-limiting input to the AI buildout.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Mon, 05 Jan 2026 07:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/1aad0c3a/1aa0d9f6.mp3" length="1337428" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>167</itunes:duration>
      <itunes:summary>For a decade, software ate the world. Now AI is making land matter again.</itunes:summary>
      <itunes:subtitle>For a decade, software ate the world. Now AI is making land matter again.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>The Real Reason Data Centers Are Moving Locations</title>
      <itunes:episode>3</itunes:episode>
      <podcast:episode>3</podcast:episode>
      <itunes:title>The Real Reason Data Centers Are Moving Locations</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">acc7c01c-c9c0-47ca-b7ff-0bb40839a99f</guid>
      <link>https://share.transistor.fm/s/eb906ac9</link>
      <description>
        <![CDATA[<p>Data centers are relocating. Not for tax incentives. Not for talent. For power.</p><p>Virginia's data center corridor consumed 26% of state electricity in 2023. Dominion Energy is struggling to keep pace. Grid stress events and connection delays stretching to years are pushing capital to Texas, Ohio, Indiana, Southeast Asia, and the Nordics.</p><p>Capital is following power geography, not network geography. The next decade of AI infrastructure will be built where grid capacity exists — not where the internet already concentrates.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>Data centers are relocating. Not for tax incentives. Not for talent. For power.</p><p>Virginia's data center corridor consumed 26% of state electricity in 2023. Dominion Energy is struggling to keep pace. Grid stress events and connection delays stretching to years are pushing capital to Texas, Ohio, Indiana, Southeast Asia, and the Nordics.</p><p>Capital is following power geography, not network geography. The next decade of AI infrastructure will be built where grid capacity exists — not where the internet already concentrates.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Fri, 02 Jan 2026 07:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/eb906ac9/7be0e57f.mp3" length="1607862" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>201</itunes:duration>
      <itunes:summary>Data centers are relocating. Not for tax incentives. Not for talent. For power.</itunes:summary>
      <itunes:subtitle>Data centers are relocating. Not for tax incentives. Not for talent. For power.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>Why Compute Is Constrained by Electricity, Not Chips</title>
      <itunes:episode>2</itunes:episode>
      <podcast:episode>2</podcast:episode>
      <itunes:title>Why Compute Is Constrained by Electricity, Not Chips</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">65931ca6-c70b-4d17-8a76-e035a8f2b370</guid>
      <link>https://share.transistor.fm/s/abba0073</link>
      <description>
        <![CDATA[<p>The GPU shortage narrative is fading. The real constraint on AI compute is electricity — and it's not even close.</p><p>Chip fabrication is scaling — TSMC, Intel, and Samsung are investing billions. But grid capacity doesn't scale like silicon. A single NVIDIA H100 draws 700 watts. AI training clusters now exceed 100 megawatts. Goldman Sachs projects 165% increase in global data center power demand by 2030.</p><p>The problem isn't that chips don't exist. The problem is you can buy the chips, but you can't plug them in.]]&gt;</p>]]>
      </description>
      <content:encoded>
        <![CDATA[<p>The GPU shortage narrative is fading. The real constraint on AI compute is electricity — and it's not even close.</p><p>Chip fabrication is scaling — TSMC, Intel, and Samsung are investing billions. But grid capacity doesn't scale like silicon. A single NVIDIA H100 draws 700 watts. AI training clusters now exceed 100 megawatts. Goldman Sachs projects 165% increase in global data center power demand by 2030.</p><p>The problem isn't that chips don't exist. The problem is you can buy the chips, but you can't plug them in.]]&gt;</p>]]>
      </content:encoded>
      <pubDate>Thu, 01 Jan 2026 07:00:00 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/abba0073/bccb85c0.mp3" length="1464923" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>183</itunes:duration>
      <itunes:summary>The GPU shortage narrative is fading. The real constraint on AI compute is electricity — and it's not even close.</itunes:summary>
      <itunes:subtitle>The GPU shortage narrative is fading. The real constraint on AI compute is electricity — and it's not even close.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
    <item>
      <title>AI Is a Power Problem, Not a Software Problem</title>
      <itunes:episode>1</itunes:episode>
      <podcast:episode>1</podcast:episode>
      <itunes:title>AI Is a Power Problem, Not a Software Problem</itunes:title>
      <itunes:episodeType>full</itunes:episodeType>
      <guid isPermaLink="false">e32f1d89-5378-4d0c-a4c6-672de38baf47</guid>
      <link>https://share.transistor.fm/s/d3940ef2</link>
      <description>
        <![CDATA[The AI boom is not constrained by chips, algorithms, or talent. It's constrained by electricity. Capital is moving accordingly.<p>In this episode, we examine why AI facilities require 50-150 kilowatts per rack versus 10-15 kilowatts for traditional computing, why developers face 5-7 year grid interconnection delays, and how Microsoft and Meta are deploying $145 billion combined in infrastructure bets on land, substations, and transmission capacity.</p><p>The companies winning this cycle won't be the ones with the best models. They'll be the ones with secured power capacity and execution timelines measured in megawatts, not parameters.</p>]]&gt;]]>
      </description>
      <content:encoded>
        <![CDATA[The AI boom is not constrained by chips, algorithms, or talent. It's constrained by electricity. Capital is moving accordingly.<p>In this episode, we examine why AI facilities require 50-150 kilowatts per rack versus 10-15 kilowatts for traditional computing, why developers face 5-7 year grid interconnection delays, and how Microsoft and Meta are deploying $145 billion combined in infrastructure bets on land, substations, and transmission capacity.</p><p>The companies winning this cycle won't be the ones with the best models. They'll be the ones with secured power capacity and execution timelines measured in megawatts, not parameters.</p>]]&gt;]]>
      </content:encoded>
      <pubDate>Tue, 30 Dec 2025 22:48:42 -0800</pubDate>
      <author>Spring Street Management Group</author>
      <enclosure url="https://media.transistor.fm/d3940ef2/73cccced.mp3" length="1448197" type="audio/mpeg"/>
      <itunes:author>Spring Street Management Group</itunes:author>
      <itunes:duration>181</itunes:duration>
      <itunes:summary>The AI boom is not constrained by chips, algorithms, or talent. It's constrained by electricity. Capital is moving accordingly.</itunes:summary>
      <itunes:subtitle>The AI boom is not constrained by chips, algorithms, or talent. It's constrained by electricity. Capital is moving accordingly.</itunes:subtitle>
      <itunes:keywords>AI infrastructure investing, data center investment, power grid AI, hyperscaler capital expenditure, data center financing, AI compute constraints, electricity demand AI, grid interconnection, data center real estate, infrastructure capital allocation</itunes:keywords>
      <itunes:explicit>No</itunes:explicit>
    </item>
  </channel>
</rss>
