Meta Taps Millions of AWS Graviton Cores to Scale Agentic AI

MetaMeta

· Updated

Meta is deploying tens of millions of AWS Graviton cores to support the high-compute demands of its next-generation agentic AI systems. The partnership signals a shift toward CPU-intensive reasoning and planning workloads as autonomous agents move to global production scale.

Meta signed an agreement with Amazon Web Services (AWS) to integrate tens of millions of AWS Graviton cores into its compute portfolio. These custom ARM-based processors handle the data orchestration required for agentic AI (autonomous systems that reason, plan, and execute multi-step tasks). This deployment makes Meta a top Graviton customer.

The move addresses a shift in infrastructure as AI evolves from chat to autonomous agents. While GPUs handle inference, the agent loop puts significant strain on CPU resources. This partnership follows Meta's recent push to diversify its hardware stack, mirroring Anthropic's massive AWS compute expansion to stabilize performance.

You can expect more stable autonomous features within Meta's ecosystem as these cores come online. The integration powers infrastructure for systems like Meta's Muse Spark reasoning model serving billions of people. This validates Graviton 5 as a primary architecture for the logic-heavy layers of agentic systems.

View the full update on about.fb.com
AI at Meta
AI at Meta
@AIatMeta
X

Today we’re announcing an agreement with Amazon Web Services to bring tens of millions of AWS Graviton cores to our compute portfolio. This partnership marks an expansion of our diversified AI infrastructure and will help scale systems behind Meta AI and agentic experiences that serve billions of people. Learn more: https://t.co/cmeATOB7Jc

13retweets217likes
View on X

Still wondering? A few quick answers below.

Meta is integrating tens of millions of AWS Graviton cores to support the infrastructure behind Meta AI and agentic experiences. These custom ARM-based CPU cores handle the complex reasoning, planning, and data processing tasks required for autonomous AI systems that serve billions of people worldwide.

While GPUs handle model inference, the agent loop is increasingly CPU-intensive. Meta follows a diversified portfolio approach to infrastructure, believing that no single chip architecture can efficiently serve every workload. AWS Graviton 5 cores are purpose-built to provide the faster data processing and greater bandwidth needed for continuous reasoning at scale.

Agentic AI refers to autonomous systems that can reason, plan, and execute complex tasks without constant human direction. These systems require significant compute power to process information and make decisions. Meta is scaling its infrastructure to ensure these agentic experiences can function efficiently for its global user base.

Meta has agreed to bring tens of millions of AWS Graviton cores into its compute portfolio. This massive deployment makes Meta one of the largest Graviton customers in the world. The agreement also includes the flexibility for Meta to expand its use of these processors as its AI capabilities and requirements grow.

Meta is specifically deploying AWS Graviton 5 cores as part of this agreement. These processors are designed by Amazon to offer the performance and efficiency required for the next generation of complex AI workloads. They provide the foundational infrastructure needed to power the reasoning and execution layers of Meta's AI systems.

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

Meta

See all AI news & updates from Meta →

Share this update