OpenAI has finalized a substantial agreement with Advanced Micro Devices (AMD), securing up to 6 gigawatts of the chipmaker’s advanced graphics processing units (GPUs) over several years, the two companies announced Monday. The multi-phase rollout—scheduled to commence in the latter half of 2026—comes as OpenAI intensifies its trillion-dollar infrastructure effort amid escalating competition in the artificial intelligence sector.
The deal cements a commitment by OpenAI to utilize AMD’s high-performance processors, deploying successive generations of GPU technology as the agreement progresses. The phased expansion will see OpenAI steadily augment its compute resources until the full 6-gigawatt milestone is achieved.
As a major component of the partnership, AMD has issued warrants permitting OpenAI to acquire as many as 160 million AMD shares at a price of $0.01 each, based on achieving certain operational and stock price milestones set out by both companies. Should all warrants be exercised, OpenAI could gain up to a 10% equity stake in AMD.
The initial share tranche will vest when AMD delivers its first batch of MI450 chips, expected in the second half of 2026. Subsequent milestones, according to executives, include AMD stock price thresholds that climb to $600 per share for the final allocation.
AMD, which currently has 1.62 billion shares outstanding and a market capitalization of $267.23 billion, saw its stock surge as much as 35% in premarket trade Monday. Executives project the OpenAI partnership could drive tens of billions of dollars in new revenue, an outlook that prompted a slight dip in shares of rival Nvidia. The latter recently announced a $100 billion collaboration with OpenAI to power next-generation AI data centers.
OpenAI CEO Sam Altman described the new AMD arrangement as “a major step in building the compute capacity needed to realize AI’s full potential,” highlighting the accelerated timeline for bringing advanced AI to a broader audience. “AMD’s leadership in high-performance chips will enable us to accelerate progress and bring the benefits of advanced AI to everyone faster,” Altman said in a statement.
Analysts, referencing data from LSEG, expect AMD to record $32.78 billion in revenue this year.
Earlier this month, Nvidia revealed a separate plan to invest as much as $100 billion in OpenAI, facilitating the rollout of massive data centers equipped with Nvidia’s processors. Under this initiative, OpenAI aims to construct 10 gigawatts’ worth of AI data center capacity, to be powered by millions of Nvidia GPUs, as highlighted by statements from both companies. This underscores a wider industry shift, where data center scale is increasingly quantified in gigawatts.
The opening phase of this Nvidia project is slated to begin with the introduction of the Vera Rubin systems, scheduled to come online in late 2026.
