Meta to add AWS Graviton cores to run agentic AI

April 24 (UPI) — Meta announced Friday that it’s partnering with Amazon Web Services to add millions of AWS Gravitron cores to Meta’s computers to boost agentic artificial intelligence.

The first deployment will begin with “tens of millions of Graviton cores, with the flexibility to expand as our AI capabilities grow,” Meta said in a press release.

“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,” Amazon Vice President Nafea Bshara said in a statement.

Bshara co-founded chipmaker Annapurna Labs, which Amazon bought in 2015.

“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,” Bshara said.

The Graviton cores will offer faster data processing and greater bandwidth, which are important for AI systems that need to continuously reason through and execute tasks.

But as Meta invests in AI, it’s also cutting staff.

Meta is also laying off 10% of its workforce — about 8,000 people worldwide and focus on AI, an employee memo said Thursday. The first round of layoffs is expected to begin May 20. Meta also said it won’t hire for 6,000 open positions.

“As we scale the infrastructure behind Meta’s AI ambitions, diversifying our compute sources is a strategic imperative,” Santosh Janardhan, Meta head of infrastructure, said in a statement. “Expanding to Graviton allows us to run the CPU-intensive workloads behind agentic AI with the performance and efficiency we need at our scale.”

The Graviton5 chip features 192 cores and a cache that’s five times larger than the previous generation, which reduces delays in how quickly those cores communicate with each other by up to 33%, Amazon said.

Graviton is built on the AWS Nitro System, which enables bare-metal instances for direct access to the hardware, while offering the same Elastic Network Adapter and Amazon Elastic Block Store devices, which allow Meta to run its own virtual machines without performance loss.