Meta has signed a deal to use millions of AWS Graviton chips to power its AI operations, Amazon announced on Friday, April 25, 2026. The agreement marks a notable shift in the type of hardware being tapped for AI work — and pulls more of Meta’s cloud spending back toward Amazon.
AWS Graviton is an ARM-based CPU (central processing unit), not a GPU (graphical processing unit). While GPUs remain the standard for training large AI models, the rise of AI agents is creating demand for a different kind of compute. Agents handle tasks like real-time reasoning, code writing, search, and coordinating multi-step processes — workloads that Amazon says its latest Graviton chip was specifically designed to address.
The deal is a win for Amazon at a pointed moment. AWS timed the announcement to coincide with the close of the Google Cloud Next conference. Meta had previously signed a six-year, $10 billion deal with Google Cloud in August of the prior year, though Meta had historically been primarily an AWS customer that also used Microsoft Azure.
Amazon’s own AI GPU, the Trainium — used for both model training and inference — was largely spoken for after Anthropic agreed earlier in April 2026 to spend $100 billion over 10 years running its workloads on AWS, with a particular focus on Trainium. In return, Amazon agreed to invest an additional $5 billion into Anthropic, bringing its total investment to $13 billion. The Graviton deal with Meta gives Amazon a separate showcase for its homegrown CPU capabilities.
The Graviton chips compete with Nvidia’s Vera CPU, which is also ARM-based and targets AI agentic workloads. A key distinction is that Nvidia sells its chips and AI systems directly to enterprises and cloud providers, including AWS, while AWS only offers access to its chips through its cloud service.
The deal reflects a broader push by Amazon to compete on price-performance in AI infrastructure. In his annual shareholder letter earlier in April 2026, Amazon CEO Andy Jassy stated that enterprises want better price-performance ratios for AI and that he intends to win deals on that basis — putting significant pressure on Amazon’s internal chip-building team to deliver.
Source: TechCrunch