AMD secured a data center contract with Meta worth double-digit billions of dollars per gigawatt, the company announced February 24, 2026. AMD stock rose 7% following the news as investors bet on the chipmaker's push into enterprise AI infrastructure.
The Meta deployment centers on AMD's Corvex confidential computing platform, which delivers near-native performance comparable to NVIDIA's HGX B200 systems. This marks the first large-scale production rollout of confidential computing infrastructure for AI workloads, addressing enterprise security requirements that have limited cloud AI adoption.
AMD's data center GPU revenue has lagged NVIDIA's dominant position in AI training chips. NVIDIA controls roughly 80% of the AI accelerator market, with data center revenue hitting $47.5 billion in Q4 2024. AMD's MI300 series GPUs launched in late 2023 but have struggled to break NVIDIA's enterprise foothold.
The Meta contract signals potential momentum for AMD in AI inference workloads, where lower precision requirements and cost sensitivity favor competition. Meta operates one of the world's largest AI infrastructure deployments, with over 600,000 GPUs planned for 2024 training clusters. Shifting even a fraction of inference workloads to AMD could generate billions in annual revenue.
Analysts will track AMD's data center segment over the next 2-4 quarters to confirm revenue conversion from the Meta contract. AMD's data center GPU revenue reached $1.9 billion in Q4 2024, up from $1.3 billion the prior quarter but still a fraction of NVIDIA's scale.
NVIDIA faces mounting competitive pressure as cloud providers and enterprises seek supply diversification. Microsoft, Google, and Amazon have all developed in-house AI chips while expanding AMD partnerships. Intel's Gaudi 3 AI accelerators also target the inference market, though adoption remains limited.
The confidential computing angle differentiates AMD's approach from pure performance competition. Corvex enables encrypted processing for sensitive AI workloads, addressing regulatory requirements in healthcare, finance, and government sectors where NVIDIA's standard offerings lack native security features.
AMD's market share in AI infrastructure will depend on execution speed, software ecosystem maturity, and sustained customer wins beyond Meta.