Nvidia and AMD have announced a joint investment of $4 billion in Meta✴ as part of an expansion of their strategic cooperation in the field of artificial intelligence. The partnership involves collaboration on the Muse family of models and the development of infrastructure for AI agents. The deal is part of a broader financial operation: Meta✴ recently announced plans to raise $95 billion through a share offering, with Nvidia, AMD, and Broadcom expressing interest in acquiring up to $5 billion worth of its shares.
Offering Structure and Financial Parameters
Meta✴'s financial operation is divided into two stages. In the first stage, the company intends to raise $35 billion through a guaranteed public offering, followed by another $60 billion through an open-market share sales program. Out of the total $4 billion that Nvidia and AMD will channel into Meta✴, Nvidia's share accounts for $3 billion, and AMD's for $1 billion. Broadcom has also confirmed its intention to participate in acquiring shares, bringing the total pool of interested investors alongside the other two companies to up to $5 billion.
Nvidia: Vera Rubin Platform and Muse Multi-Agent System
Nvidia intends not only to invest $3 billion in Meta✴ but also to expand the use of the Vera Rubin platform and Vera processors in its partner's infrastructure. The company will deepen cooperation in developing the comprehensive multi-agent AI system Muse, which was unveiled in September 2026. Nvidia CEO Jensen Huang stated in the context of the deal that AI agents represent a new stage in the evolution of computing technologies, where software understands the user's goals and performs work on their behalf. Thus, Nvidia's investment is both financial and technological, cementing its chips as the foundation for agentic computing within the Meta✴ ecosystem.
AMD: Helios Racks and EPYC Processors Already in Action
AMD is investing $1 billion in Meta✴ and announced that its Helios server racks are already operating within the company's infrastructure, while EPYC processors are being used to power the Muse system. According to AMD CEO Lisa Su, the partners are developing gigawatt-scale computing infrastructure and working together on new generations of hardware, software, and systems for Meta✴'s future models and services. The mention of gigawatt scale indicates that these are data centers consuming energy comparable to the output of large power plants, which is typical for modern AI clusters.
Market Reaction and Investment Activity in 2026
Against the backdrop of the investment reports, AMD shares rose by 3% on Tuesday, Nvidia stock remained virtually unchanged, and Meta✴ shares fell by 1%. At the same time, both investing companies are actively increasing their investments in artificial intelligence market participants. Since the beginning of 2026, Nvidia has participated in 121 investment rounds compared to 13 in 2022. AMD has also expanded its investment activity: in juillet, the company announced plans to invest up to $5 billion in Anthropic, and the total number of funding rounds involving it in 2026 reached 25 compared to two in 2022. These figures highlight the semiconductor vendors' systemic strategy to secure positions in the AI ecosystem through direct investments in key consumers of their computing platforms.
Context: Meta✴'s Proprietary AI Chips
It is important to note that alongside securing investments from Nvidia and AMD, Meta✴ is developing its own proprietary AI chips. It was previously reported that the company expects to begin production of its AI chip as early as September 2026 to reduce its dependence on Nvidia. Thus, the $4 billion investment deal creates a model of mutual dependency: Meta✴ receives funding and access to advanced computing platforms, while Nvidia and AMD solidify their chips as the standard for the infrastructure of the world's largest AI model developer, simultaneously diversifying their investment portfolios amidst the race for dominance in agentic AI.
