OpenAI’s Big AMD Deal

What It Covers

OpenAI secures massive chip deals with AMD and NVIDIA totaling over one trillion dollars in infrastructure commitments, while the AI arms race drives compute, memory, and energy costs skyward across the entire technology industry.

Key Questions Answered

• •OpenAI-AMD Partnership Structure: OpenAI commits to six gigawatts of AMD Instinct GPUs over multiple years, potentially acquiring 10% AMD ownership through warrants tied to deployment milestones, reducing dependence on single vendor NVIDIA while spending billions on diversified chip supply.
• •AI Infrastructure Economics: Memory and storage costs surge 120% year-over-year as OpenAI's Stargate project locks in 900,000 DRAM wafers monthly—40% of global output—creating decade-long supply constraints that increase prices for consumer electronics from Raspberry Pi to gaming consoles and laptops.
• •Compute as Competitive Moat: Elon Musk spends over 18 billion dollars acquiring 300,000 NVIDIA chips for Colossus two, burning through 13 billion dollars in 2025 alone, demonstrating belief that raw compute power determines AI market winners, making speed and chip volume critical competitive advantages.
• •Small Nuclear Reactor Economics: Small modular reactors cost 180 dollars per megawatt hour versus 133 dollars for conventional nuclear and 126 dollars for natural gas, with wind and solar one-third the price, potentially making SMRs economically unviable despite 32 gigawatts of signed agreements.