- Meta commits to tens of thousands of hosted Graviton cores
- The agreement includes infrastructure, network, power and management layers
- Graviton5 is designed for continuous processing and multi-tasking
Meta has signed an agreement to deploy tens of thousands of AWS Graviton Arm cores, making it one of the largest Graviton customers in the world.
The deal marks a major expansion of the long-standing partnership between Meta and AWS, but with a critical difference: Meta isn’t just buying chips; it is to buy the entire infrastructure around them. This is a wholesale deal, not a hardware purchase.
“As we scale the infrastructure behind Meta’s AI ambitions, diversifying our computing resources is a strategic imperative,” said Santosh Janardhan, head of infrastructure at Meta.
The article continues below
Why agent AI is driving massive demand for CPUs
While GPUs remain essential for training large AI models, the rise of agent AI is creating a massive demand for CPU-intensive workloads.
Agent systems perform real-time reasoning, code generation, search operations, and orchestration of multistage tasks, all of which rely heavily on CPU power.
Graviton5 is purpose-built for these workloads, offering faster processing and greater bandwidth than mainstream alternatives.
The chip has 192 Arm Neoverse V3 cores, 600 megabytes of total cache and support for DDR5 8800 memory and PCIe Gen6.
Meta will use these chips to support its AI tools that require handling billions of interactions while coordinating complex multi-step agent workflows.
This deal can be seen as a major blow to AMD and Intel, two companies that have traditionally dominated the large-scale infrastructure CPU market.
Meta is not only buying tens of thousands of Graviton cores, but also the power, data center space, networking and AWS management tools around those cores.
This means that Meta chooses AWS’s vertically integrated infrastructure rather than buying off-the-shelf chips and integrating them into its own data centers.
How the Graviton infrastructure works
The Graviton5 is built on 3nm chip technology, a manufacturing process that produces smaller, more efficient processors.
It features a Nitro system that enables bare metal instances while providing familiar networking and storage devices.
This allows Meta to run its own virtual machines without compromising performance.
AWS designs its chips from scratch and manages the entire process from chip design to server architecture.
Therefore, it can optimize performance and efficiency in ways that ordinary processors cannot match.
Meta appreciates such a capability because it means the chips can be tuned to achieve specific performance levels.
As Meta’s AI capabilities grow, deployment will likely extend beyond the initial tens of thousands of cores.
The agreement between Meta and AWS is clearly important, but it is important to distinguish between the purchase of infrastructure and the purchase of chips.
Meta does not take delivery of Graviton processors for installation in its own data centers; it rents AWS-hosted computing capacity at massive scale.
“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, predicts and scales efficiently to billions of people worldwide,” said Nafea Bshara, Vice President and Distinguished Engineer, Amazon.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds.
