Meta has announced a major expansion of its artificial intelligence infrastructure through a new agreement with Amazon Web Services that will bring tens of millions of AWS Graviton cores into its compute portfolio, making the company one of the largest Graviton customers in the world. The move signals a significant shift in how Meta is building the next generation of AI systems, particularly as it deepens its focus on agentic AI.
Meta Bets Big on CPU Power for Agentic AI
While much of the AI industry has focused heavily on GPUs for training large language models, Meta says the rise of agentic AI is changing what modern infrastructure needs to look like. Agentic AI systems — which are designed to reason, plan and autonomously execute complex tasks — place growing pressure on CPU resources, especially for workloads such as real-time reasoning, search, code generation and task orchestration.
To support this shift, Meta will begin deploying tens of millions of AWS Graviton5 cores, with the flexibility to scale further as its AI ambitions grow. These processors are purpose-built to handle CPU-intensive workloads and offer the speed and bandwidth required for complex AI systems operating at global scale.
A Strategic Shift in AI Infrastructure
The deal reflects a broader strategic pivot at Meta as it diversifies the compute systems powering its AI roadmap. Rather than relying on a single chip architecture, Meta says it is building a more flexible infrastructure model that combines in-house hardware, custom data centre investments and external cloud partnerships.
Santosh Janardhan, Head of Infrastructure at Meta, said the agreement is rooted in the company’s long-term infrastructure strategy.
“As we scale the infrastructure behind Meta’s AI ambitions, diversifying our compute sources is a strategic imperative,” said Janardhan. “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.”
AWS Strengthens Its Position in the AI Race
For AWS, the agreement marks a major win in the rapidly intensifying AI infrastructure race. The deal not only expands its long-standing relationship with Meta, but also positions AWS and its custom Graviton silicon as a serious force in the future of AI compute.
Nafea Bshara, Vice President and Distinguished Engineer at Amazon, said the partnership is about more than chips.
“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,” said Bshara.
Building AI at Global Scale
The agreement underscores how the AI arms race is evolving beyond GPUs, with CPUs becoming increasingly critical in powering the reasoning and orchestration layers of next-generation AI systems. For Meta, the partnership with AWS is less about buying hardware and more about securing the infrastructure needed to scale AI services efficiently across billions of users.
