What's happening
Meta Platforms signed a multi-billion dollar agreement with Amazon Web Services on April 24, 2026, to deploy tens of millions of Graviton CPU cores for agentic AI workloads. The partnership makes Meta one of the world's largest customers for AWS's Graviton processors, which will power AI infrastructure serving billions of users across Meta's platforms. Each Graviton5 processor contains 192 cores and features 5x larger cache than previous generations, up to 33% reduced communication delays, and up to 25% better performance. The deal represents a strategic shift toward CPU-intensive workloads for complex AI tasks like orchestration and reasoning, complementing the industry's focus on GPU-based training and inference.
Why it matters for markets
The multi-billion dollar commitment underscores the expanding role of CPUs in AI infrastructure beyond traditional GPU dominance, particularly for agentic AI applications that require complex reasoning and orchestration capabilities. For Amazon, the deal strengthens AWS's position in the $200 billion cloud computing market and validates its custom silicon strategy against competitors like Intel and AMD. Meta's diversification into Graviton processors supports its massive AI infrastructure spending, which has exceeded $20 billion annually, while potentially reducing costs compared to traditional x86 processors. The partnership reinforces both companies' positioning in the rapidly growing AI infrastructure market, with Meta's $1.71 trillion market capitalization reflecting investor confidence in its AI investments and Amazon's $2.84 trillion valuation supported by AWS's continued expansion.
Sectors and assets to watch
Cloud computing infrastructure providers face increased competitive pressure as hyperscalers like Amazon expand custom silicon offerings. Traditional CPU manufacturers including Intel and Advanced Micro Devices may see market share pressure in data center processors as cloud providers develop proprietary alternatives. Other major cloud platforms including Microsoft Azure and Google Cloud Platform will likely accelerate their own custom processor development to compete with AWS's Graviton advantage. Meta's commitment to tens of millions of cores signals potential opportunities for semiconductor equipment manufacturers and foundries supporting custom silicon production at scale.
What to watch next
Monitor Meta's quarterly infrastructure spending disclosures for evidence of the Graviton deployment's financial impact and timeline for the multi-billion dollar commitment. Track AWS's custom silicon revenue growth and customer adoption metrics in upcoming Amazon earnings reports. Watch for competitive responses from Microsoft, Google, and other cloud providers regarding their own custom processor initiatives and major customer wins in the AI infrastructure space.