AMD Lands 6-Gigawatt GPU Deal With OpenAI, Its Largest Yet

AMD and OpenAI have signed a definitive agreement to deploy 6 gigawatts of AMD Instinct GPUs across multiple chip generations, with the first 1-gigawatt installation of Instinct MI450 series hardware scheduled to start in the second half of 2026. The deal, announced October 6, 2025, in Santa Clara, is the most significant commercial validation to date of AMD's push into AI accelerators, a market where it has trailed Nvidia.

The scale is the story. Six gigawatts of compute is enough capacity to power a large city's worth of AI training and inference workloads, and OpenAI has committed to it under a multi-year, multi-generation agreement that makes AMD a core strategic compute partner. OpenAI will deploy AMD technology "starting with the AMD Instinct MI450 series and rack-scale AI solutions, and extending to future generations," the companies said.

A warrant tied to milestones

To align incentives, AMD has issued OpenAI a warrant for up to 160 million shares of AMD common stock. The warrant vests in tranches tied to concrete outcomes. The first tranche vests with the initial 1-gigawatt deployment. Additional tranches vest as OpenAI's purchases scale up to the full 6 gigawatts. Vesting is also conditioned on AMD hitting certain share-price targets and on OpenAI achieving the technical and commercial milestones required to deploy AMD hardware at scale.

That structure is unusual for a chip supply deal. It gives OpenAI direct upside in AMD's stock price while penalizing both sides if the deployment stalls. It also signals that AMD is willing to pay heavily — in equity-linked value — to lock in a marquee customer.

The financial stakes for AMD are large. Jean Hu, AMD's EVP, CFO and treasurer, said the partnership "is expected to deliver tens of billions of dollars in revenue for AMD while accelerating OpenAI's AI infrastructure buildout." She added that the agreement "creates significant strategic alignment and shareholder value for both AMD and OpenAI and is expected to be highly accretive to AMD's non-GAAP earnings-per-share."

Why it matters

OpenAI's compute demand has outpaced what any single supplier can provide, and the company has moved to line up multiple hardware partners for its next-generation infrastructure. For AMD, landing a 6-gigawatt commitment from the highest-profile AI lab answers the central question hanging over its data center GPU business: whether its Instinct accelerators can win deployments at the largest scale, not just in niche configurations.

The two companies frame the deal as the continuation of a working relationship that already spans two product generations. The collaboration "began with the MI300X and continued with the MI350X series," according to the announcement. Under the new agreement, the companies will share technical expertise to optimize their product roadmaps, a process OpenAI has used with hardware partners to shape silicon and software for its specific workloads.

AMD's statement points to its "strong leadership in high-performance computing systems" and OpenAI's "pioneering research and advancements in generative AI" as the basis for pairing the two companies at what it calls "this important and pivotal time for AI."

What the executives said

The announcement carries quotes from four senior leaders, and their emphasis differs in ways that map onto each company's priorities.

Dr. Lisa Su, AMD's chair and CEO, focused on scale and mutual benefit: "We are thrilled to partner with OpenAI to deliver AI compute at massive scale. This partnership brings the best of AMD and OpenAI together to create a true win-win enabling the world's most ambitious AI buildout and advancing the entire AI ecosystem."

Sam Altman, OpenAI's co-founder and CEO, framed the deal around capacity: "This partnership is a major step in building the compute capacity needed to realize AI's full potential. AMD's leadership in high-performance chips will enable us to accelerate progress and bring the benefits of advanced AI to everyone faster."

Greg Brockman, OpenAI's co-founder and president, stressed the full-stack nature of the collaboration: "Building the future of AI requires deep collaboration across every layer of the stack. Working alongside AMD will allow us to scale to deliver AI tools that benefit people everywhere."

Hu's statement, as CFO, anchored the deal in financial terms — the tens of billions in expected revenue and the accretion to non-GAAP EPS.

The timeline

The first deliverable arrives in the second half of 2026, when OpenAI begins deploying the AMD Instinct MI450 Series at 1 gigawatt. That gives AMD roughly a year to ramp MI450 production and rack-scale systems at volumes the Instinct line has not yet seen. Everything beyond that first gigawatt — the remaining 5 gigawatts — follows as purchases scale, with warrant tranches unlocking along the way.

The MI450 succeeds the MI300X and MI350X in the lineage the two companies have worked on together. The companies describe the partnership as extending to "future generations" of Instinct hardware, meaning the 6-gigawatt figure is not a single-order quantity but a rolling commitment tied to AMD's roadmap execution.

What to watch

The warrant's vesting conditions give observers a precise scorecard. If AMD's share price reaches the specified targets and OpenAI's deployments hit the technical and commercial milestones, the tranches vest and the 6 gigawatts materialize on schedule. Any slippage shows up directly in the vesting schedule. The second half of 2026 delivery date for the first gigawatt of MI450 hardware is the first checkpoint.