AMD secured a multi-billion dollar data center contract from Meta on February 24, 2026, worth double-digit billions per gigawatt. The stock rose 7% following the announcement, adding roughly $14 billion in market value.
The contract represents Meta's first large-scale deployment of AMD chips for AI workloads, breaking NVIDIA's near-monopoly on enterprise AI infrastructure. Meta has relied almost exclusively on NVIDIA's H100 and A100 GPUs for training and inference across its platforms including Facebook, Instagram, and WhatsApp.
Corvex deployment technology enabled the shift by achieving near-native performance for confidential computing on NVIDIA HGX B200 systems. This allows enterprises to run AMD chips alongside existing NVIDIA infrastructure without performance penalties, lowering the technical barriers to switching suppliers.
AMD's data center segment generated $3.5 billion in Q4 2025, up 69% year-over-year. The Meta contract could add $2-3 billion annually to this segment over the next 2-4 quarters, based on typical hyperscaler deployment timelines.
NVIDIA still controls roughly 80% of the AI accelerator market, with 2025 data center revenue reaching $96 billion. But Meta's move signals potential vulnerability. Amazon, Microsoft, and Google have all developed custom AI chips to reduce NVIDIA dependence, though none have scaled them beyond internal workloads.
AMD trades at 28x forward earnings versus NVIDIA's 32x, despite slower growth rates. Analysts at KeyBanc raised AMD's price target to $185 from $160, citing the Meta deal as proof AMD can compete in high-performance AI inference.
Intel announced its Gaudi 3 AI chip in December 2025 but has secured no major hyperscaler contracts. The company's data center GPU revenue fell 8% in Q4 2025 to $4.0 billion.
Investor focus now shifts to AMD's Q1 2026 earnings on April 29, where management will likely provide Meta deployment details and updated data center guidance. Wall Street expects 15-20% upward revision to full-year 2026 data center revenue estimates if Meta's rollout accelerates.