Chipmaker AMD is expanding its artificial intelligence (AI) infrastructure partnership with Meta, including a planned $1-billion equity investment, in a move that mirrors the investment strategy pursued by rival Nvidia before the controversial departure of its longtime CEO Jensen Huang.
AMD said it has begun testing its Helios rack-scale computing systems at Meta, with broader deployment expected as the social media company expands the computing infrastructure supporting its AI models and services.
The agreement covers processors, graphics processing units (GPUs), networking equipment, and software, with both companies working toward deploying AI infrastructure at gigawatt scale.
AMD’s planned equity investment in Meta also reflects a broader trend among chipmakers of providing financial backing to companies that purchase or deploy their computing technologies.
Nvidia has pursued a similar strategy through investments in AI startups and infrastructure providers, including CoreWeave, OpenAI, and Nebius.
Such arrangements, sometimes described as “circular financing,” have drawn scrutiny because chipmakers invest in companies that subsequently spend substantial amounts on their hardware and services.
The concerns have intensified following Huang’s abrupt resignation as Nvidia CEO on November 12 amid reports about his personal conduct.
His departure has raised questions about the future of Nvidia’s financing arrangements and whether the company will continue investing heavily in AI startups and infrastructure providers that rely on its GPUs.
AMD, meanwhile, has been pursuing its own investment-backed partnerships. In July, the company announced a commitment to invest up to $5 billion in AI developer Anthropic as part of an agreement involving the deployment of AMD Instinct GPUs.
Its latest investment in Meta extends that approach to one of the world’s largest technology companies.
“Meta is a strategic partner for AMD, and we have built this partnership through years of deep technical collaboration – from EPYC processors powering Meta’s infrastructure to Instinct GPUs and open software supporting its AI workloads,” said Lisa Su, chair and CEO of AMD.
“With Helios now running at Meta, we are taking that partnership to a new scale in the superintelligence era. Together, we are building and deploying infrastructure at gigawatt scale and shaping future generations of silicon, software and systems to enable Meta’s most ambitious superintelligence models and services.”
Meta already uses millions of AMD EPYC processors across several generations and is a lead customer for its sixth-generation EPYC CPUs.
The processors also support the computing environments behind Muse, Meta’s personal AI agent.
Meta also uses AMD Instinct GPUs for AI training, ranking, and recommendation systems. The company plans to deploy a customized Instinct GPU optimized for its workloads.
The two companies are collaborating on future processor and GPU architectures, including AMD’s Verano processors and Instinct MI500 series.
Their software collaboration includes improvements to PyTorch, Triton, and AMD’s ROCm platform, as well as support for Meta’s Muse Spark Open and Muse Glimmer AI models.
“AMD has been a key partner in building out the compute that’s running our frontier models,” said Mark Zuckerberg, founder and CEO of Meta.
“We’re making big investments in the computing power that will enable us to bring personal superintelligence to everyone, and AMD plays an important role in that.”
The expanding relationship highlights how competition between AMD and Nvidia is extending beyond chip performance and pricing to include financing arrangements, long-term infrastructure commitments, and investments in companies developing AI services.
AMD did not disclose a timetable for its full Helios deployment at Meta or additional financial terms for the planned $1-billion investment.
