Meta buys tens of millions of AWS Graviton 5 processor cores from Amazon

2026-04-24 · Source: The Decoder · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Robotics & Autonomous Systems · Depth: Intermediate, quick

Summary

Meta is acquiring tens of millions of AWS Graviton 5 processor cores from Amazon, establishing itself as one of the largest Graviton customers globally. These ARM-based CPU cores are intended to power Meta's agentic AI systems, which require significant processing for task planning and execution, complementing GPUs used for AI model training. This purchase aligns with Meta's broader strategy of investing in ARM-based processors, following its deployment of Nvidia's Grace CPUs in February and the co-development of the "AGI CPU" with ARM in March. The agreement with AWS is scalable, allowing Meta to potentially transition workloads from Graviton 5 to its new ARM chips in the future. Specific financial terms or contract duration were not disclosed.

Key takeaway

For CTOs and VP of Engineering evaluating infrastructure for advanced AI, your strategy should prioritize ARM-based CPUs like Graviton 5 for agentic AI orchestration, even as GPUs handle training. This move by Meta signals a critical shift towards distributed, CPU-intensive AI workloads, suggesting a need to diversify beyond GPU-centric planning to optimize cost and performance for complex AI systems.

Key insights

Meta is heavily investing in ARM-based CPUs, including AWS Graviton 5, to power agentic AI systems.

Principles

• CPUs are critical for agentic AI orchestration.
• ARM architecture gains data center traction.

In practice

• Utilize ARM CPUs for AI system coordination.
• Diversify CPU suppliers for AI infrastructure.

Topics

• AWS Graviton 5
• Agentic AI Systems
• ARM Processors
• CPU Orchestration
• Data Center Chips

Best for: CTO, VP of Engineering/Data, MLOps Engineer, AI Engineer, AI Architect, Director of AI/ML

Related on AIssential

• In another wild turn for AI chips, Meta signs deal for millions of Amazon AI CPUs — AWS secured a significant Meta deal for Graviton CPUs, intensifying cloud and custom AI chip competition.
• Meta to Deploy “Millions” of Nvidia Processors — AI investment is driving significant shifts in chip deployment, software development, and market dynamics across multiple sectors.
• Arm debuts AGI CPU as it moves into silicon manufacturing for AI — Arm's AGI CPU targets agentic AI infrastructure, expanding Arm into direct silicon manufacturing.
• Google's TPU 8 Is A Direct Attack On NVIDIA - And It Rewrites AI Infrastructure Forever — Hyperscalers are shifting to specialized AI chips and vertically integrated stacks for the agentic era.
• A new version of OpenAI’s Codex is powered by a new dedicated chip — Dedicated AI hardware accelerates lightweight models for real-time developer collaboration and rapid iteration.
• Guest post: OpenClaw Is Powerful – But the Agent Tax Is Slowing It Down* — Agentic AI workflows demand specialized hardware to overcome high latency and token costs, enabling faster, more thorough execution.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Decoder.