AMD has secured a significant multi-year chip supply agreement with OpenAI, projected to generate tens of billions in revenue for the chipmaker. This collaboration is set to considerably enhance AMD's position within the rapidly expanding artificial intelligence sector. The core of this agreement involves AMD furnishing OpenAI with a colossal 6 gigawatts of compute capacity, equivalent to the power required for approximately 4.5 million households. This immense capacity will be delivered through multiple iterations of AMD's advanced Instinct GPUs, commencing with the Instinct MI450 model. OpenAI anticipates receiving the initial gigawatt of this capacity in the latter half of 2026, coinciding with the MI450 GPU's scheduled rollout.
A key facet of this strategic alliance is AMD's assertion that its MI450 series will surpass Nvidia's comparable offerings, such as the NVIDIA Rubin CPX, in performance. This superior capability is attributed to a combination of hardware and software enhancements, many of which will be developed in close consultation with OpenAI. Currently, AMD's MI355X and MI300X series GPUs are already demonstrating strong performance in AI inference for large language models, largely due to their substantial memory capacity and bandwidth. Furthermore, the agreement includes a provision for OpenAI to acquire up to 160 million shares of AMD stock, representing a 10% stake in the company. The vesting of these shares is linked to both the deployment of compute capacity and the achievement of specific AMD stock price milestones, with the final tranche contingent on the stock reaching $600 per share. Following the announcement, AMD's stock saw a nearly 35% increase, opening at $222.24 after closing at $164.67.
This partnership underscores OpenAI's aggressive efforts to forge numerous chip alliances as it races to construct a robust AI infrastructure. These endeavors include the development of five new Stargate data centers in collaboration with Oracle and SoftBank, designed to achieve a combined capacity of 7 gigawatts. Just last month, OpenAI secured an investment of up to $100 billion from Nvidia, which will also provide at least 10 gigawatts of compute capacity. Additionally, OpenAI entered a $10 billion agreement with Broadcom for the development and production of bespoke AI chips, and expanded its Stargate initiative. Recent weeks have also seen OpenAI finalize deals with Samsung Electronics and SK Hynix to procure DRAM memory chips for the Stargate project, with plans to establish data centers in South Korea.
The collaboration between AMD and OpenAI is a testament to the accelerating pace of innovation in the AI landscape, emphasizing the critical need for advanced computing power. This strategic partnership not only positions both companies at the forefront of AI development but also signals a collective commitment to pushing the boundaries of what artificial intelligence can achieve. By combining cutting-edge hardware with pioneering AI research, they are paving the way for a future where AI's transformative potential can be fully realized, driving progress and fostering an ecosystem of unprecedented technological advancement.
