Meta to use 6GW of AMD GPUs, days after expanded Nvidia AI chip deal
Lisa Su, president and chief executive officer of Advanced Micro Devices Inc., showcases the AMD Instinct MI455X GPU at the 2026 CES event in Las Vegas on January 5, 2026.
Bloomberg | Bloomberg | Getty Images
a week later Meta determined to use millions Nvidia’s The social media company, which is working with processors to power its AI expansion, has signed a new giant chip deal this time. Advanced Micro Devices.
The multi-year deal with AMD includes the deployment of up to 6 gigawatts of graphics processing units for the company’s AI data centers and the use of AI-optimized central processing units, or CPUs, Meta said Tuesday. The first shipments of MI450 GPUs for AMD’s Helios rack-scale servers will begin later this year.
AMD CEO Lisa Su said in a statement that her company “delivers a high-performance, energy-efficient infrastructure optimized for Meta workloads, accelerating one of the industry’s largest AI deployments and placing AMD at the center of the global AI fabric.”
The deal also includes a performance-based guarantee for Meta to acquire 160 million AMD shares, approximately 10% of the company. The first tranche goes into effect when the first 1GW of Instinct GPUs ships. Other tranches were also vacated due to Meta purchasing up to 6GW.
Vesting is also tied to stock price thresholds for AMD and technical and commercial milestones for Meta.
AMD struck a similar deal with OpenAI in October, marking it as a viable second option for AI giants and hyperscalers.
This deal also guaranteed OpenAI to purchase 160 million AMD shares and was subject to distribution and stock price criteria.
AMD and Meta one day stock chart.
In its earnings report last month, Meta pledged to spend up to $135 billion in capital spending this year as it tries to keep up with OpenAI and Anthropic, as well as megacap peers, in the global AI race. Overall, Meta has plans for 30 data centers, 26 of which are in the US
Tuesday’s announcement is a critical development for AMD, which has fallen far behind Nvidia in the AI chip market. Nvidia is currently the world’s largest publicly traded company with a valuation of $4.66 trillion and controls nearly 90% of the market, while AMD is valued at $320 billion.
“Meta is in a unique position to control the entire stack, and they can use whoever’s computation they want,” said chip analyst Ben Bajarin of Creative Strategies. “This is just a punctuation point that we are computationally constrained and deals will be made across the board.”
Terms of the deal were not specified, but Bajarin, who has knowledge of the deal, estimates it is worth tens of billions of dollars over at least four years because “six gigawatts will take quite some time to deploy,” he said.
Bajarin said a key part of the deal, and how it differs from Meta’s deal with Nvidia, is that the initial distribution includes customized GPUs.
“We have no indication that Nvidia is doing this,” Bajarin said. “It’s a good way for them to win some of these deals.”
AMD’s Helios represents the first large-scale competition against Nvidia’s Grace Blackwell systems, which have seen increasing demand since their launch in 2024.
Nvidia is scheduled to report quarterly earnings on Wednesday, and analysts expect revenue to rise 68% from a year earlier to $66 billion, according to LSEG. Earlier this month, AMD reported that its fourth-quarter sales rose 34% to $10.27 billion.
Meta has been a long-time customer of AMD and Nvidia. Meta is also developing in-house processors and was in talks with Meta, according to a report published late last year in The Information. Google About the deployment of the search company’s tensor processing units to Meta data centers in 2027.
WRISTWATCH: How did AMD become the chip giant and eventually catch up with Intel?
