OpenAI Commits to 6GW GPU Deployment in Major Partnership with AMD
OpenAI and AMD Announce Partnership with 6GW GPU Deployment Deal
OpenAI has formed a strategic partnership with AMD to deploy 6 gigawatts (GW) of the latter’s Instinct GPUs, establishing AMD as a core compute provider for the AI company. The news suggests a significant expansion of OpenAI’s infrastructure strategy.
The agreement specifies a multi-year deployment schedule, with the initial 1GW of AMD Instinct MI450 GPUs set to begin installation in the second half of 2026. Purchases will then scale up to the full 6GW capacity.
Central to the deal’s strategic alignment, AMD has granted OpenAI a warrant for up to 160 million shares of common stock. This structured warrant, which could result in OpenAI acquiring approximately 10% ownership of AMD if fully exercised, is tied to the completion of deployment milestones, AMD’s share-price targets, and OpenAI achieving the technical capability to use the chips at scale. AMD CFO Jean Hu stated the agreement is “expected to deliver tens of billions of dollars in revenue” and will be “highly accretive” to AMD’s non-GAAP earnings-per-share.
In statements regarding the collaboration, AMD CEO Dr. Lisa Su emphasised that the partnership will enable the “world’s most ambitious AI buildout,” while OpenAI CEO Sam Altman noted that AMD’s chips will allow them to “accelerate progress” and bring the benefits of advanced AI to a broader audience. OpenAI President Greg Brockman highlighted the need for “deep collaboration across every layer of the stack” to scale their AI tools.
This agreement with AMD is the latest in a series of strategic infrastructure deals by OpenAI, which has recently secured similar partnerships with technology giants like NVIDIA and Samsung in its consistent effort to secure the necessary compute capacity for advanced AI development.
