Key Highlights
- Meta has entered into a massive multi-year agreement with Amazon to deploy AWS Graviton CPU cores for AI infrastructure.
- The partnership encompasses deployment of tens of millions of Graviton cores spanning a three to five-year period.
- Meta shares advanced 0.6% while Amazon stock climbed 1.4% in premarket activity after the announcement.
- Amazon joins Meta’s diverse chip vendor ecosystem alongside Nvidia, Broadcom, and AMD.
- Meta now ranks among the top five Graviton clients for AWS.
Meta has finalized a multibillion-dollar partnership with Amazon that will provide the social media giant access to tens of millions of Amazon Web Services’ Graviton CPU cores for its artificial intelligence agent initiatives.
According to Nafea Bshara, an Amazon VP who co-founded Annapurna Labs — AWS’s proprietary chip division — the arrangement spans three to five years. The majority of these Graviton cores will be positioned across U.S. data centers.
Following the disclosure, Amazon shares rose 1.4% during premarket hours, with Meta experiencing a 0.6% uptick.
This partnership centers exclusively on AWS Graviton central processing units rather than graphics processing units. While GPUs dominated AI infrastructure conversations until recently, CPUs are experiencing renewed relevance.
The emergence of AI agents has repositioned CPUs as critical components. These processors manage distinct operations and channel workloads to GPUs, creating a synergistic relationship between both chip categories across diverse AI applications.
CPUs are equally vital during the “post-training” stage of large language model development, where base models undergo refinement for targeted objectives.
Meta selected Amazon’s Graviton5, built on 3-nanometer architecture, based on its cost-effectiveness and performance metrics. “Meta has access to so many options from the supply side. But they chose Graviton5,” Bshara noted.
Meta Maintains Multi-Vendor Chip Approach
This Amazon partnership expands Meta’s existing supplier network that features Nvidia, Broadcom, AMD, and Arm Holdings. The company has emphasized its commitment to avoiding dependence on any single chip platform.
“No single chip architecture can efficiently serve every computational task,” Meta stated in its official announcement.
Initial rollout will involve tens of millions of Graviton cores, with potential expansion tied to Meta’s evolving AI requirements.
Meta Doubles Down on AI While Trimming Workforce
Meta’s commitment to artificial intelligence continues accelerating. The company completed its acquisition of AI startup Manus for over $2 billion last December. Manus specializes in developing AI agents for sophisticated operations, creating additional CPU requirements.
To offset AI infrastructure expenses, Meta revealed Thursday plans to reduce its workforce by approximately 10% — eliminating roughly 8,000 positions — effective in May.
Meta introduced Muse Spark, its first new AI model in twelve months, earlier this month, with additional releases planned for the near future.
While Meta and AWS have collaborated since approximately 2016, previous engagements primarily involved cloud infrastructure, the Bedrock platform, and GPU cluster leasing. This agreement represents a significant shift toward custom chip partnerships.
For AWS, securing Meta as a Graviton client represents substantial validation. Just days earlier, Amazon disclosed a $5 billion commitment to Anthropic, which similarly incorporates tens of millions of Graviton CPU cores.
AWS began developing proprietary chips before 2018, when the inaugural Graviton processor launched using Arm architecture.
Bshara confirmed that this partnership elevates Meta into AWS’s top five Graviton customers.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
