Meta strikes agreements with Nvidia and AMD for AI chips and equipment

Meta Platforms' multi-year supply agreements with Nvidia and Advanced Micro Devices, announced on 2026-03-05, represent a critical strategic move to secure the vast AI compute capacity required for its "massive energy and data demands." This deal sees Meta acquiring "hundreds of thousands" of AI GPUs, including Nvidia's H100 and AMD's Instinct MI300X, for its operating stage data centers. The commercial logic for Meta centers on diversifying its supply chain for these essential components, mitigating reliance on a single vendor for its foundational AI infrastructure. Nvidia benefits from Meta's continued commitment to its H100 GPUs, reinforcing its market leadership in accelerated computing. Simultaneously, Advanced Micro Devices gains a substantial order for its Instinct MI300X, expanding its market share in the competitive AI chip landscape. This dual-vendor strategy ensures Meta's infrastructure can scale with future AI requirements. For Meta Platforms, a global technology company, these multi-year agreements directly support its infrastructure strategy, which is "primarily driven by the massive energy and data demands" of its operations. Nvidia, specializing in "accelerated computing, particularly in AI," secures a substantial order for its H100 GPUs, reinforcing its market leadership in data center technology. Advanced Micro Devices, a global semiconductor company, significantly expands its presence in the "infrastructure industry" with the inclusion of its Instinct MI300X GPUs in Meta's procurement. This announced supply agreement, involving "hundreds of thousands" of units, demonstrates the intense competition among major technology companies to secure advanced AI hardware.