Oracle is dropping a staggering $40 billion on Nvidia’s GB200 AI chips to build what could be one of the world’s largest AI data centers in Abilene, Texas. The 875-acre facility will house approximately 400,000 “superchips” delivering mind-boggling computational power (we’re talking quintillions of calculations per second, folks). By partnering with OpenAI, SoftBank, and others, Oracle aims to dominate the AI infrastructure game while reducing OpenAI’s dependence on Microsoft. The full scope of this Texas-sized gamble might surprise you.
Forty billion dollars. That’s the eye-watering sum Oracle is dropping on Nvidia’s GB200 AI chips for a massive data center project in Abilene, Texas. To put that in perspective, you could buy 40 billion McChickens or fund NASA for nearly two years—but Larry Ellison clearly has more ambitious plans for his cash.
The purchase includes roughly 400,000 of Nvidia’s latest “superchips,” which will power what’s set to become one of the world’s largest AI data centers. Each rack delivers over 1 exaflop of training performance—that’s a quintillion calculations per second, for those keeping score at home. Remember when a gigaflop sounded impressive? Those were simpler times.
This Texas-sized investment is part of the broader Stargate initiative, a project potentially worth up to $500 billion. Oracle isn’t going solo on this digital frontier—they’ve assembled an Avengers-level team including OpenAI, SoftBank, Microsoft, and Arm to create an AI powerhouse.
The facility’s 1.2 gigawatts of computing power would make Doc Brown jealous. “Great Scott!” indeed. By mid-2026, this computational behemoth will span eight buildings, growing faster than tech startups burn through venture capital. The impressive Abilene site spans 875 acres of Texas land, providing ample room for future expansion.
What’s Oracle’s endgame? They’ll lease this massive computational capacity to OpenAI under a 15-year agreement. It’s a clever pivot that both diversifies Oracle’s business model and helps OpenAI reduce its dependency on Microsoft’s cloud resources. With AI technologies expected to generate $15.7 trillion in revenue by 2030, Oracle’s strategic position could pay off enormously. The partnership represents a strategic shift for OpenAI, which had previously relied on Microsoft for nearly $14 billion in cloud credits. Talk about a high-stakes tech love triangle.
The financial engineering behind this deal is equally impressive. JPMorgan arranged $9.6 billion in debt, with another $5 billion in equity from Crusoe and Blue Owl Capital. Wall Street clearly believes the AI hype is real.
With this investment, Oracle isn’t just buying chips—it’s purchasing a front-row seat in the AI revolution. The question remains whether this $40 billion gamble will look visionary or foolhardy a decade from now. Place your bets, folks.
