Add MobiGyaan as a preferred source on GoogleMeta has expanded its infrastructure partnership with Amazon Web Services (AWS), announcing a large-scale deployment of Graviton processors to support next-generation AI systems. The agreement includes tens of millions of Graviton cores, with scope to scale further as Meta builds out its future AI infrastructure.
The partnership signals a shift in AI infrastructure strategy:
- GPUs remain critical for model training
- CPUs are becoming essential for agentic AI workloads
Meta is optimizing for a hybrid approach where different chips handle different AI tasks efficiently.
Rise of Agentic AI workloads
Meta is using Graviton processors for CPU-intensive AI tasks, including real-time reasoning, code generation, and search and orchestration across large systems. These workloads involve continuous, large-scale computation, making CPUs more efficient than GPUs for certain operations.
Graviton5 and AWS Nitro Architecture
Meta’s deployment will rely heavily on Graviton5, AWS’s latest processor.
Key specs:
- 192-core architecture
- 5x larger cache vs previous gen
- Up to 33% lower inter-core latency
Supporting technologies:
- AWS Nitro System → near bare-metal performance
- Elastic Fabric Adapter (EFA) → low-latency, high-bandwidth communication
This setup is designed for distributed AI workloads at a massive scale.
Graviton5 is built on 3nm process technology, enabling up to 25% better performance vs the previous generation, improved energy efficiency, and lower operational cost per workload. This aligns with Meta’s push toward scalable yet sustainable AI infrastructure.
Nafea Bshara, Vice President and Distinguished Engineer, Amazon, said, “This isn’t just about chips; it’s about giving customers the infrastructure foundation, as well as data and inference services, to build AI that understands, anticipates, and scales efficiently to billions of people worldwide. Meta’s expanded partnership, deploying tens of millions of Graviton cores, shows what happens when you combine purpose-built silicon with the full AWS AI stack to power the next generation of agentic AI.”
Santosh Janardhan, Head of Infrastructure, Meta, said, “As we scale the infrastructure behind Meta’s AI ambitions, diversifying our compute sources is a strategic imperative. AWS has been a trusted cloud partner for years, and expanding to Graviton allows us to run the CPU-intensive workloads behind agentic AI with the performance and efficiency we need at our scale.”
