The 150 to 8 Problem
The xAI Memphis Colossus supercomputer is the concrete case study behind Jon Najarian’s Infinite Power pitch, and the numbers tell the whole story. The facility needed 150 megawatts to run. The local utility could deliver about eight. That is not a rounding error; it is a 95% shortfall at the exact moment a company with enormous capital needed power immediately. What happened next is why the AI energy trade exists at all. Our full Infinite Power teardown covers the presentation this all comes from.
The 122-Day Fix
Rather than wait in a multi-year interconnection queue, xAI bypassed the grid entirely. It trucked in roughly 35 gas-turbine “power pods,” units 91 feet long and weighing 200,000 pounds each, and had them generating more than 400 megawatts in about 122 days. That speed is the headline insight: a grid that takes years to deliver power lost the deal to a fleet of trailers that did it in four months.
The same playbook is now scaling at Southaven, Mississippi, and every major data center operator is watching, because the bottleneck the Colossus hit is about to be hit by everyone building AI capacity. There is more on this in our Data Center Electricity: The Bottleneck Explained article.
Why This Is a Signal, Not an Anecdote
The Memphis buildout matters because it revealed the true shape of the AI energy opportunity. The constraint is not chips, not capital, but grid interconnection, and the companies that solve that constraint fastest capture the growth. That is why the pitch argues the suppliers, not the front-page names, will be the biggest winners. Vertiv Holdings (VRT), the power and cooling infrastructure supplier up roughly 1,200% from its SPAC debut, and Bloom Energy (BE), the fuel-cell name cited as proof the market pays up for AI power, are the precedent names. We broke out the behind-the-meter power mechanics separately, and the Jon Najarian profile covers who is making the call. We break down that angle in our Bloom Energy Stock News for 2026: What Matters piece.
The Honest Caveats
The Colossus story is real and checkable, but the investment conclusions drawn from it deserve scrutiny. A company that derived 96% of its 2024 pod revenue from xAI is a single-customer bet, however compelling the contracted pipeline looks. And the “replace the entire grid” framing overreaches; most of the multi-trillion-dollar grid spend is utilities’ wires and transformers, not on-site turbines for data centers. Najarian discloses the concentration risk himself, which is more than most presenters do, but the risk remains the risk.
The Bottom Line
The xAI Memphis Colossus is the clearest single example of the AI energy bottleneck and the fastest solution to it. It is not a secret, but it is the event that made the supplier layer investable, and it is the reason a gas-turbine leasing company is suddenly the center of a newsletter pitch. The hardware story is settled; what is still open is which specific suppliers actually capture the growth, and at what price.
NewsletterVetter is an independent publication. We receive compensation from some of the services we review through affiliate links. Nothing on this site is investment advice. Always do your own research.