Meta has signed a deal with Amazon Web Services (AWS) to deploy “tens of millions” of Graviton5 cores in support of agentic AI workloads.

The exact financial terms or timeline were not shared, but in response to a DCD request for comment, AWS said the agreement represented a "multibillion-dollar deal over several years."

In a separate, pre-prepared statement, Meta said the deal will make it one of the largest Graviton customers in the world.

Following the initial deployment, the social media giant said it will have the ability to expand its procurement to meet its AI compute needs.

The announcement comes two weeks after Amazon president and CEO Andy Jassy revealed in a letter to shareholders that two large AWS customers had asked to purchase every single Graviton instance capacity available this year so far.

“We can’t agree to these requests given other customers’ needs, but it gives you an idea of the demand,“ he wrote.

AWS launched its Graviton5 CPUs in December 2025, claiming the hardware has an "efficient design" that reduces inter-core communication latency by up to 33 percent. Graviton5 also has a five times larger cache than the previous generation, enabling customers to scale up workloads while reducing infrastructure costs, the company said at the time.

Each chip has 192 cores, meaning if Meta were to deploy exactly 10 million cores, that would represent just over 52,000 chips. In response to a DCD request for clarification regarding the size of the deployment, AWS said it would not be disclosing the exact number of chips.

“As we scale the infrastructure behind Meta’s AI ambitions, diversifying our compute sources is a strategic imperative,” said Santosh Janardhan, head of infrastructure, Meta. “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.”

Nafea Bshara, VP and distinguished engineer, Amazon, added: “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.”

CPUs are processors with a limited number of cores, making them best suited for serial computing tasks. GPUs, by comparison, have thousands of smaller cores and are used for parallel computing tasks, such as AI and high-performance computing.

Demand for CPUs has increased significantly of late, as companies seek out hardware that can support agentic AI workloads.

This is the latest CPU-focused deal Meta has inked in recent months. In February 2026, the company signed an agreement with Nvidia that represents the first “large-scale” Nvidia Grace-only deployment, with Meta also committing to an additional “large-scale deployment” of Nvidia’s forthcoming Vera CPUs in 2027.

In March, it was revealed that Meta had partnered with Arm for the chip company’s first-ever data center CPU, designed to support “agentic AI infrastructure.”

Earlier this month, Meta announced it was partnering with Broadcom for the development of multiple generations of the hyperscaler’s MTIA (Meta Training and Inference Accelerator) chips, designed to support generative AI inference workloads.