<div id="readability-page-1"><div><p>//php echo do_shortcode('[responsivevoice_button voice=&#34;US English Male&#34; buttontext=&#34;Listen to Post&#34;]') ?&gt;</p> <p>Google’s parent Alphabet has reached a definitive agreement to acquire renewable energy developer Intersect Power for $4.75 billion, a transaction that signals a structural transformation in how Silicon Valley intends to power the AI era.</p><p>The deal, which includes taking on Intersect Power’s debt, makes Alphabet the first major tech company to bring electricity generation in-house. By owning a power utility, Google can secure energy for its data centers directly.</p><p>This acquisition marks a departure from the industry’s decade-long standard of signing Power Purchase Agreements, where companies contract for energy from third-party developers. Instead, Google is taking ownership of a 3.6-gigawatt pipeline of late-stage solar and wind projects, along with 3.1 gigawatt-hours of battery storage.</p> <p>“Intersect has always been focused on bringing innovation to the industry, and we look forward to accelerating at scale as part of Google,” said Sheldon Kimber, founder and CEO of Intersect. “Modern infrastructure is the linchpin of American competitiveness in AI. We share Google’s conviction that energy innovation and community investment are the pillars of what must come next.”</p>  <p>Google’s new strategy comes as the U.S. electricity sector faces its biggest supply challenges in decades. These are caused by more EVs, increased domestic manufacturing, and the rapid growth of AI computing.</p> <h3>Gigawatt AI problem</h3><p>The main reason for the deal is how much energy generative AI uses. For example, a ChatGPT query needs about 2.9 watt-hours, almost 10× more than a regular Google search. As Alphabet plans to double its capital spending to $185 billion in 2026 to support these technologies, finding enough electricity has become the main challenge.</p><p>“The energy system, I would say, globally, is no longer fit for purpose, for serving the demands of AI,” said Amanda Peterson Corio, Google’s global head of data center energy. “What we’re trying to do is really think proactively.”</p><p>By buying Intersect, Google gets control of a developer with projects ready to start and, importantly, spots already secured in interconnection queues. In many U.S. markets, it now takes five to seven years for new power projects to connect to the grid. By owning these projects, Google can focus on serving its data centers in places like Texas, instead of waiting for the general market to add capacity.</p><p>Oliver Kerr, managing director for North America at Aurora Energy Research, noted the strategic logic of the move. “Just as Google brought in-house some of the elements around chip design to reduce reliance on Nvidia, I think it makes sense that they’re also bringing the power-generation piece in-house.”</p><h3>Divergent strategies among hyperscalers</h3><p>While Google is buying a developer, its competitors are taking other approaches to the same energy challenge. Microsoft has focused on financial strategies and bringing back nuclear power. In September 2024, Microsoft signed a 20-year deal with Constellation Energy to restart Unit 1 of the Three Mile Island nuclear plant, which will add 835 megawatts of carbon-free power to the grid by 2028.</p><p>At the same time, Microsoft started a $30 billion global AI infrastructure investment partnership with BlackRock to fund energy projects. This helps deal with the reluctance of traditional lenders to finance such infrastructure.</p><p>Amazon is taking a different approach by buying distressed assets. Last month, it bought the 1.2-gigawatt Sunstone solar-plus-storage project from a bankrupt developer for $83 million. Like Google, Amazon wants to control energy assets directly, but it is focusing on buying individual projects instead of an entire development company.</p><h3>Nuclear renaissance and small reactors</h3><p>Besides renewables, the industry is investing heavily in nuclear energy to provide constant power, something wind and solar cannot do without large battery storage. Tech companies are betting on small modular reactors (SMRs), which can be built faster than traditional nuclear plants.</p><p><strong>Comparative analysis of hyperscaler nuclear strategies</strong></p><figure><table><thead><tr><th><strong>Company</strong></th><th><strong>Partner</strong></th><th><strong>Technology</strong></th><th><strong>Target Capacity</strong></th><th><strong>Target Date</strong></th></tr><tr><th><strong>Microsoft</strong></th><th>Constellation Energy</th><th>Restarted PWR (Three Mile Island)</th><th>835 MW</th><th>2028</th></tr><tr><th><strong>Google</strong></th><th>Kairos Power</th><th>Fluoride Salt-Cooled SMR</th><th>500 MW</th><th>2030</th></tr><tr><th><strong>Amazon</strong></th><th>X-energy</th><th>High-Temp Gas SMR (Xe-100)</th><th>5 GW (Goal)</th><th>~2030+</th></tr><tr><th><strong>Oracle</strong></th><th>Unknown</th><th>SMRs (3 Units)</th><th>~1 GW</th><th>Unconfirmed</th></tr></thead></table></figure><p>Source: Market Reports and Corporate Filings</p><p>Prior to the Intersect acquisition, Google had signed an agreement with Kairos Power to install 500 MW of SMR capacity by 2035, using a new salt-cooled design. Amazon has invested $500 million in X-energy to help deploy high-temperature gas-cooled reactors.</p><p>However, the timelines for these technologies remain long. “There’s no free lunch, there’s no magic wand that makes all the problems go away,” said Tim Latimer, CEO of Fervo Energy, a geothermal startup backed by Google. “That’s something I think Google internalized a lot sooner than the rest of their peers.”</p><h3>Economics of power integration</h3><p>Owning energy assets gives Google clear benefits but also brings new risks. By controlling the developer, Google can use the “Energy Park” model, placing data centers next to power sources. This setup can help avoid transmission congestion charges and maintain power quality. Sheldon Kimber stressed the need for this scale. “You don’t even get a seat at the table here without being a big, big player in physical infrastructure.”</p><p>However, this strategy also brings regulatory attention. The Federal Energy Regulatory Commission (FERC) recently rejected a plan to place an Amazon data center at a nuclear plant in Pennsylvania, saying it could shift transmission costs to others. This shows that tech companies can’t easily avoid the financial rules of the larger power grid.</p><h3>Regulatory headwinds and consumer protection</h3><p>The idea of tech giants controlling power generation has worried utilities and consumer advocates. In Ohio, regulators approved a “Large Load Tariff” that makes data centers pay for 85% of their contracted capacity, even if they don’t use it all. This protects residential customers from paying for upgrades that might not be needed.</p><p>Brad Smith, Microsoft’s president, tried to address these worries in January 2026 with a “Community First” pledge. The company promised to pay all extra costs for its electricity use and to give up property tax discounts in the communities where it operates.</p><p><strong>Emerging regulatory frameworks for large loads</strong></p><figure><table><tbody><tr><td><strong>Provision</strong></td><td><strong>Detail</strong></td><td><strong>Economic Implication</strong></td></tr><tr><td><strong>Take-or-Pay</strong></td><td>Mandatory payment for ~85-90% of capacity</td><td>Prevents “phantom load” bookings</td></tr><tr><td><strong>Exit Fees</strong></td><td>Multi-year demand charges upon termination</td><td>Disincentivizes stranding assets</td></tr><tr><td><strong>Credit Assurance</strong></td><td>Collateral required for non-investment grade</td><td>Protects utility balance sheets</td></tr><tr><td><strong>“Downward Pressure”</strong></td><td>Incremental revenue must lower residential rates</td><td>Mandates cross-subsidization benefit</td></tr></tbody></table></figure><p>Source: AEP Ohio Tariff Filings and Georgia Power</p><h3>Future outlook</h3><p>As we approach 2030, the line between tech companies and power utilities will become less clear. The “hypergrid” model predicts that data centers will shift from being passive users to active grid participants, using the same AI tools to reduce computing during peak demand. Google has already shown this by using “carbon-intelligent computing” to pause non-urgent tasks when the grid is under stress, acting as a demand response resource.</p><p>Furthermore, Alphabet’s CEO Sundar Pichai said that even with these large investments, supply will have trouble keeping up, and he “still expects demand from its DeepMind research lab and from customers to outstrip the new computing power it can bring online”.</p><p>Buying Intersect Power shows that for the biggest tech companies, energy is now something to manage, not just buy. The key question for the AI era is whether this approach can overcome the limits of the aging U.S. electric grid.</p><hr/><h5>See also:</h5><p>Silicon Carbide Growth Pivots Around AI and Energy</p><p>Delivering Energy for Abundant AI Will Require New Playbook</p><p>Solving AI’s Power Struggle</p> </div></div>

GGLS

GGLL

<p>Google's parent company, Alphabet, has agreed to acquire renewable energy developer Intersect Power for $4.75 billion, marking a significant shift in how tech companies manage energy for AI. This acquisition allows Google to control a 3.6-gigawatt pipeline of solar and wind projects, addressing the growing energy demands of AI technologies. Unlike competitors, Google is bringing power generation in-house, aiming to secure energy for its data centers directly. This strategic move comes amid challenges in the U.S. electricity sector, driven by increased energy consumption from AI and electric vehicles.</p>

Google Bets Big on Power Generation to Fuel AI