AI News Feed
Market watch
X tests animated arm pointing users toward promoted postsBank of America Warns Customers Not to Share Banking Passwords With AI AssistantsMozilla Adds AI Form Filling and Website Monitoring to Firefox Smart WindowResearchers tire of OpenAI gatekeeping as company unveils math, physics and computer science claimsiOS 27.2 Gives Apple Music a Black-and-White Icon and Drops Its Signature RedSamsung to require parental approval for some apps on children’s Galaxy phones
X tests animated arm pointing users toward promoted postsBank of America Warns Customers Not to Share Banking Passwords With AI AssistantsMozilla Adds AI Form Filling and Website Monitoring to Firefox Smart WindowResearchers tire of OpenAI gatekeeping as company unveils math, physics and computer science claimsiOS 27.2 Gives Apple Music a Black-and-White Icon and Drops Its Signature RedSamsung to require parental approval for some apps on children’s Galaxy phones
AI Chips & Compute

Nvidia, AMD to Buy $4 Billion of Meta Shares as Chip Partnerships Expand

Nvidia and AMD will buy $4 billion of Meta shares in expanded chip deals tied to a stock sale of up to $95 billion.

AI News Feed · 2026-10-28 07:35

Nvidia is committing $3 billion and AMD $1 billion. Broadcom Inc. is also participating in the stock sale. Meta did not specify Broadcom's investment amount but said the three chipmakers together plan to invest $5 billion, suggesting Broadcom will buy $1 billion.

Under the expanded Nvidia partnership, Meta plans to broaden its use of Rubin graphics processing units and Vera central processing units, which are often deployed together in Nvidia-designed DGX Vera Rubin NVL72 appliances. Each system includes 72 GPUs and 36 CPUs. Meta also plans to buy more Nvidia networking gear and expand its use of the Spectrum-X Ethernet product family, which includes SuperNICs and switches. Nvidia says Spectrum-X checks GPU traffic for anomalies at sub-microsecond intervals to address connectivity bottlenecks. Meta will also use BlueField-4 accelerators to speed networking, storage and cybersecurity tasks.

Nvidia will help Meta improve the performance and cost efficiency of its Muse large language model family and the company's consumer AI agent. Meta also plans to explore collaboration opportunities related to Nvidia's OpenShell Reference Architecture, a collection of technical resources for running AI agents in isolated sandboxes designed to mitigate cybersecurity risks. "NVIDIA and Meta are expanding our partnership to optimize the computing infrastructure end to end for capable, efficient and secure AI agents, so Muse can empower billions of people to achieve more and pursue greater ambitions," Nvidia CEO Jensen Huang said.

AMD, which is investing $1 billion, said Meta has adopted its Helios rack design. Helios contains 72 GPUs, like Nvidia's DGX Vera Rubin NVL72, and is based on AMD's latest Instinct MI450 graphics card architecture, Epyc CPUs and Pensando data processing units, AMD's counterpart to BlueField-4. AMD said Meta has deployed an initial set of Helios racks and is testing them, with additional systems planned in the near future. Later, Meta will adopt a custom AI accelerator based on AMD's Instinct architecture, and AMD will use technical data from the partnership to improve future chip and rack designs. "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," AMD CEO Lisa Su said.