- Meta Commits to Six Gigawatts of AMD Instinct GPU Hardware
- Initial deployment of one gigawatt is planned for the second half of 2026
- Custom MI450 silicon designed specifically for Meta AI workloads
AMD and Meta agreed to deploy up to 6 GW of AMD Instinct GPUs over multiple generations, marking one of the largest infrastructure commitments disclosed in the AI industry.
The scale of the plan suggests a significant expansion of Meta’s internal computing capacity as demand for training and running advanced AI models continues to grow.
The first phase of shipments is expected to begin in the second half of 2026 and will support an initial deployment of 1 GW.
First implementation schedule takes shape
That launch will be based on a custom GPU derived from AMD’s MI450 architecture and integrated within the company’s rack-scale Helios platform.
The systems will also use sixth-generation EPYC processors, codenamed Venice, along with ROCm software.
The deal builds on an existing relationship where Meta has already deployed a large number of previous EPYC CPUs and Instinct GPUs in its global data center.
As AI workloads increase in complexity, the balance between CPU and GPU has become more critical, particularly for orchestration, scheduling, and efficiency in massive clusters.
Beyond the hardware supply agreement, the companies are co-designing a custom GPU optimized for Meta’s AI workloads.
The MI450-based custom design is expected to support large-scale AI training and inference tasks across all Meta services.
Meta says AI infrastructure is critical to providing services to billions of users, including recommendation systems, generative tools, and future applications that rely on intensive computing.
“We are excited to form a long-term partnership with AMD to implement efficient inference computing and deliver personal superintelligence,” said Mark Zuckerberg, founder and CEO of Meta.
“This is an important step for Meta as we diversify our computing. I look forward to AMD being an important partner for many years to come.”
For AMD, the deal deepens its presence in hyperscale data centers at a time when competition in AI accelerators remains intense.
“We are proud to expand our strategic partnership with Meta as it pushes the boundaries of AI at an unprecedented scale,” said Dr. Lisa Su, president and CEO of AMD.
“This multi-year, multi-generation collaboration between Instinct GPUs, EPYC CPUs and rack-scale AI systems aligns our roadmaps to deliver 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 global AI development.”
The partnership also includes a performance-based guarantee that allows Meta to acquire up to 160 million AMD shares, with rights tied to shipping and stock milestones.
The first tranche is awarded after the initial gigawatt of GPU shipments, while additional tranches depend on increasing purchases until the total commitment of 6 GW is reached.
These structures link business success directly to equity ownership, aligning incentives as infrastructure expands.
“We expect this partnership to drive substantial revenue growth over several years and contribute to our non-GAAP earnings per share,” said Jean Hu, executive vice president, chief financial officer and treasurer of AMD.
“The performance-based structure also closely aligns AMD and Meta around execution and long-term value creation.”
Follow TechRadar on Google News and add us as a preferred source to receive news, reviews and opinions from our experts in your feeds. Be sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form and receive regular updates from us on WhatsApp also.
