Meta Platforms has announced a significant partnership with Advanced Micro Devices (AMD) for artificial intelligence (AI) chips, reportedly valued at over $100 billion. This multi-year, multi-generation collaboration aims to rapidly scale Meta’s AI infrastructure and accelerate the development and deployment of advanced AI models.
Under the terms of the agreement, AMD will supply Meta with up to 6 gigawatts of AMD Instinct computing power. This includes AMD’s custom MI450 GPUs and EPYC CPUs, with the first gigawatt deployment anticipated to begin in the second half of 2026. A notable component of the deal is a performance-based stock incentive, which could see Meta awarded up to 160 million AMD shares, representing approximately 10% of AMD’s total stock, as specific milestones tied to Instinct GPU shipments are achieved.
According to AMD Chair and CEO Lisa Su, the partnership with Meta signifies a “multi-year, multi-generation collaboration.” Meta CEO Mark Zuckerberg also affirmed the long-term nature of the partnership, stating, “This is an important step for Meta as we diversify our compute. I expect AMD to be an important partner for many years to come.” Zuckerberg has previously hinted at Meta’s ambitious AI goals, including plans to deliver “personal superintelligence.”
This substantial deal with AMD follows closely on the heels of another major strategic partnership Meta announced with Nvidia. Meta plans to utilize millions of Nvidia’s current Blackwell and upcoming Rubin GPUs, along with standalone Nvidia Grace CPUs, in its production data centers, a move expected to significantly enhance performance-per-watt. This indicates Meta’s diversified approach to securing critical AI hardware from multiple vendors.
In addition to these external partnerships, Meta is also actively developing its own in-house AI chips. The company introduced its next generation of Meta Training and Inference Accelerator (MTIA) chips in April 2024, designed to improve performance for its AI workloads, particularly in ranking and recommendation models. This initiative aims to reduce reliance on external suppliers and optimize chip performance for Meta’s unique AI requirements.
Meta’s commitment to AI infrastructure is evident in its overall investment strategy. The company reportedly plans to invest between $60 billion and $65 billion in 2025 for AI and other infrastructure, a significant increase from its estimated spending in 2024. Furthermore, Meta expects to spend at least $600 billion on U.S. data centers and related infrastructure by 2028, underscoring its massive long-term investment in scaling AI capabilities.