OpenAI just recently made a huge deal with AMD. Fittingly, that partnership kicked off with a massive cloud contract — one worth 6 gigawatts of compute capacity, enough to supercharge OpenAI’s AI infrastructure. This partnership has the potential to be worth tens of billions of dollars. It’s a testament to OpenAI’s ambitious plan to construct their own data centers and grow their compute muscle to keep up with the expansive needs of AI technologies.
OpenAI CEO Sam Altman expressed enthusiasm about the partnership, stating that it represents “a major step in building the compute capacity to realize AI’s full potential.” During the first year of this partnership, OpenAI is expected to onboard its first gigawatt of compute capacity. This milestone will be reached in the second half of 2026, just as new chips for large scale AI applications are coming online.
Artificial intelligence wonderkind OpenAI is working with AMD. Beyond that, it’s signed deals with Samsung Electronics and SK Hynix for delivery of the DRAM memory chips required for its Stargate project. Under this initiative, the South Korean government announced plans to build five data centers. Collectively, they hope to reach an installed capacity of 7 gigawatts. These moves signal that OpenAI is serious about lining up chip partners, further shoring up the company’s infrastructure in the cutthroat AI race.
Unlike some previous deals, the agreement with AMD has a staged approach. That first tranche will only vest once we deploy our initial 1 gigawatt. Further tranches will be unlocked as OpenAI continues to procure up to the full 6 gw. Considering this strategic move alongside rising competition and what can be viewed as a race to keep pace with technology development, this move certainly makes sense.
In making the announcement, Dr. Lisa Su, chair and CEO of AMD, called it the company’s pleasure to engage in this partnership.
“We are thrilled to partner with OpenAI to deliver AI compute at massive scale.” – Dr. Lisa Su
OpenAI’s AI largess runs much deeper than its partnership with AMD. It’s recently locked down major, guaranteed investments from Nvidia, which has pledged to invest up to $100 billion and provide at least 10 gigawatts of capacity. OpenAI signed a $10 billion deal with Broadcom to develop and manufacture custom AI chips tailored for their needs.
The Stargate initiative has grown exponentially. In collaboration with partners Oracle and SoftBank, OpenAI is taking innovation in artificial intelligence to new heights. The company is actively working on innovative solutions that require substantial computational power, underscoring the urgency in establishing a robust infrastructure.
A TechCrunch event scheduled for October 27-29, 2025, in San Francisco will likely shed more light on these developments, offering insights into the future of AI technologies and infrastructure. Rebecca Bellan Rebecca is a senior reporter at TechCrunch. She writes about business and policy developments around AI, and she’ll be writing here about these transformative technologies.
