BREAKING: Elon Musk just warned that 15 gigawatts of AI compute...

These 10 companies are positioned to benefit most:
1. $VST - Vistra
The power producer already feeding the machines. Q2 EBITDA jumped 30% to $1.77 billion, and they just unveiled a $1 billion Helix data center partnership with KKR and Nvidia. When compute is stranded without electricity, the company selling electrons sets the price.
Proof of what desperate buyers pay: Amazon signed a 17 year agreement for 1.92 gigawatts of Talen's Pennsylvania nuclear power, then expanded the relationship. When a hyperscaler locks up two decades of output from one plant, that tells you everything about where power prices are headed.
The largest nuclear fleet in America, now the heart of the AI power story. The Microsoft deal is funding the Crane reactor restart, they signed 920 megawatts of fresh power deals last month, and nuclear expansion for AI demand is officially underway. Scarce, clean, always-on power is exactly what hyperscalers are bidding for.
The grid queue skip. Fuel cells generate power onsite, no interconnection wait required. Bloom just posted its first billion dollar quarter at $1.065 billion in revenue and raised full year guidance, with Oracle deploying their fuel cells to power AI data centers that can't wait for the grid. When the shortage hits, Bloom sells the workaround.
The punchline of the whole thread. While everyone above sells power to desperate buyers, IREN already secured theirs: contracted capacity and grid connections locked in years early, which is a big reason Microsoft handed them a $9.7 billion AI cloud contract. In a power-short 2027, energized capacity is the scarcest asset in tech. IREN is sitting on it.
If you found this useful, follow me @JasonL_Capital for more breakdowns on the supply chains powering the next decade.
Get 25% off for life with code OPTIONS25 - only 50 spots at this price:
whop.com/wheelhouse-opt…









