26. February 2026
Amd Secures 100 Billion Deal With Meta As Ai Chip Market Heats Up
Meta Platforms has agreed to purchase up to 6 gigawatts of advanced computing power from chipmaker Advanced Micro Devices (AMD), a multiyear pact valued at over $100 billion. This strategic partnership marks a significant milestone for AMD as it seeks to challenge Nvidia’s dominance in the booming AI chip market.
According to reports, Meta will utilize AMD’s Helios architecture and MI450 GPUs to develop cutting-edge AI technologies and power its next-generation AI infrastructure. The agreement also includes the possibility of Meta acquiring up to 10% of AMD’s stock through performance-based warrants, contingent on GPU shipments and financial milestones.
AMD’s CEO Lisa Su described the deal as “one of the industry’s largest AI deployments” that places AMD at the center of the global AI buildout. Su emphasized the importance of this partnership, stating that it aligns the companies’ roadmaps to deliver high-performance, energy-efficient infrastructure optimized for Meta’s workloads.
The agreement comes amidst a surge in demand for AI chips and a spate of high-value chip supply agreements. Last week, Meta announced its commitment to using millions of Nvidia processors as it ramps up its AI expansion plans. The company aims to deploy tens of gigawatts of data center computing power this decade, with ambitions to increase that number significantly over time.
Meta’s CEO Mark Zuckerberg stated that the deal represents an “important step” in diversifying the company’s compute resources. By partnering with AMD, Meta seeks to reduce its dependence on any single supplier and ensure a stable supply chain for its AI needs.
AMD is betting big on this partnership, which it believes will provide meaningful traction against Nvidia’s stranglehold on the AI chip market. The deal follows a similar arrangement with OpenAI in October, which also included stock warrants linked to delivery and performance targets.
The custom approach taken by AMD in developing chips specifically for Meta’s workloads is crucial. Meta wants these chips optimized for “inference” rather than just training, a subtle yet important distinction as AI moves from research labs into the hands of billions of users across Facebook, Instagram, and WhatsApp.
This partnership extends beyond GPUs, with Meta set to be a lead customer for AMD’s forthcoming 6th Gen EPYC processors, codenamed “Venice,” and a specialized variant called “Verano” designed for workload-specific optimizations. As part of the agreement, Meta will have the opportunity to own a significant stake in AMD through performance-based warrants.
The deal is structured to vest incrementally as specific milestones are achieved, with the first tranche unlocking with the initial 1 gigawatt of chip shipments. However, there’s another catch: Meta would only receive the final tranche of shares once AMD’s stock hits $600, according to The Wall Street Journal.
AMD has issued Meta warrants to buy up to 160 million shares of its stock, roughly 10% of the company, at a symbolic price of just one cent per share. This is a creative financing approach that mirrors the deal AMD struck with OpenAI last October, signaling that the chipmaker is leaning into innovative partnerships to lock in key customers.
This partnership highlights AMD’s strategy to diversify its customer base and reduce dependence on any single supplier. By partnering with Meta, AMD seeks to tap into the social media giant’s vast resources and expertise, positioning itself as a major player in the rapidly evolving AI landscape.
The deal also underscores the growing importance of custom silicon solutions in the AI chip market. As companies like Meta and OpenAI push the boundaries of AI research, they require tailored solutions that optimize performance, energy efficiency, and scalability.
Industry analysts note that this partnership represents a significant milestone for AMD, which has long sought to challenge Nvidia’s dominance in the AI chip market. The deal showcases AMD’s commitment to innovation and its ability to adapt to changing customer needs.
Ben Bajarin, CEO and principal analyst at Creative Strategies, says the move signals that Nvidia recognizes the growing demand for general-purpose CPU architectures in data centers, driven by agentic AI software. However, he notes that CPUs are still just one component of advanced AI hardware systems, with GPUs playing a more prominent role.
As the AI arms race continues to gain momentum, this partnership highlights the evolving landscape of chip supply agreements and the growing importance of custom solutions in the industry. With Meta’s commitment to deploying millions of Nvidia processors and AMD’s agreement to provide up to 6 gigawatts of AI computing power, the stage is set for a high-stakes battle between these two tech giants.
The deal also raises questions about the use of chips as collateral in supply agreements, as seen in Nvidia’s recent partnership with CoreWeave. According to The Information, AMD has entered into a similar arrangement with a cloud startup buying its processors, creating a circular deal that could have significant implications for the industry.
As AMD and Meta continue to push the boundaries of AI research and deployment, it remains to be seen how this partnership will shape the future of the AI chip market. However, one thing is clear: this deal marks an important step forward in the evolution of custom silicon solutions and their role in powering the next generation of AI systems.
In a rapidly changing landscape where the demand for advanced computing power continues to grow, partnerships like this one between Meta and AMD demonstrate the importance of collaboration and innovation in shaping the future of technology. As the industry continues to navigate the complexities of AI development and deployment, it will be fascinating to see how these partnerships evolve and shape the market.