Santa Clara, 06 October 2025 — OpenAI has entered into a landmark multiyear agreement with AMD to deploy expansive GPU infrastructure, a deal worth tens of billions of dollars that has sent AMD’s stock soaring and signals a serious shift in the AI computing landscape.
Under the agreement, OpenAI plans to roll out 6 gigawatts of AMD-powered compute capacity in the coming years, beginning with the first gigawatt deployment in late 2026. To align their interests, AMD has granted OpenAI a warrant to acquire up to 160 million shares, equivalent to about 10% of AMD’s outstanding shares, under favorable terms and milestone vesting.
Market Reaction & Strategic Stakes
AMD shares jumped as much as 38%, marking one of the biggest single-day rallies for the company in nearly a decade. Investors see this as validation of AMD’s ambition to challenge Nvidia’s dominance in AI infrastructure. The deal also signals to cloud providers, data center operators, and chip suppliers that OpenAI is diversifying its hardware relationships.
For AMD, this deal represents a transformational inflection, if successfully executed, it could substantially expand its AI compute business, raise margins, and deepen its presence in next-generation infrastructure. For OpenAI, the move reduces reliance on a single chip provider and secures scale and alignment in its compute roadmap.
Challenges & What to Watch
- Execution risk: Deploying gigawatt-scale compute infrastructure requires matching investment in cooling, power delivery, interconnects, and data center design. Any misstep in integration or scaling could hurt timelines and margins.
- Competitive backlash: Nvidia and others are unlikely to cede ground quietly. Expect intensified competition on chip performance, pricing, and ecosystem support.
- Shareholder mechanics: The 10% warrant stake gives OpenAI significant non-voting equity upside; how AMD manages dilution, governance, and alignment will be under scrutiny.
- Supply chain & geopolitics: AMD’s reliance on foundries like TSMC and supply networks across Asia means that regional policies, trade restrictions, or export controls could influence execution.
For Asia, the deal holds particular importance: many chip fabrication, memory, and assembly supply chains span Korea, Taiwan, Japan, Malaysia, and other nations. This injection of demand could flow downstream, benefiting regional component suppliers and strengthening infrastructure investment in AI computing hubs.
