Meta is expanding its AI capabilities with Amazon’s energy-efficient Graviton chips while continuing investments across multiple hardware partners.
Meta Platforms has signed a multibillion-dollar agreement with Amazon to deploy AWS Graviton processors across its global data center network, marking another significant step in the technology industry’s accelerating artificial intelligence investment cycle.
Under the deal, Meta will roll out Graviton CPU chips across 32 data centers over the next three years. While financial terms were not disclosed, the agreement highlights the scale of spending required to support growing AI workloads as companies compete to expand computing capacity.
Amazon’s Graviton processors are ARM-based central processing units designed for cloud workloads operating on Amazon Elastic Compute Cloud (EC2). Unlike graphics processing units (GPUs), which are typically used to train AI models, CPUs play a key role after deployment by handling inference, automation, and compute-intensive tasks required by AI agents and large-scale applications.
Meta said diversifying its computing infrastructure is a strategic priority as it scales systems supporting advanced AI services. The company expects Graviton chips to improve operational efficiency while helping manage energy consumption across massive data center operations. Amazon Web Services noted that the latest generation of Graviton processors delivers higher core density, larger cache capacity, and improved communication between cores, enabling faster processing with better performance efficiency.
The partnership comes amid a surge in AI-related spending across the technology sector. Meta has recently expanded cloud collaborations with Google Cloud while continuing major investments in AI hardware supplied by Nvidia and Advanced Micro Devices.
At the same time, Meta is developing its own custom AI silicon in collaboration with Broadcom, reflecting a multi-vendor strategy aimed at ensuring long-term access to computing resources needed to support next-generation AI platforms and large-scale intelligent systems.
