AMD has indicated interest in purchasing a portion of Meta’s proposed $95 billion stock sale, which is intended to fund Meta’s artificial intelligence infrastructure. This potential investment follows just weeks after AMD spent $8.2 billion on an AI software acquisition.

This interest, first reported by Analytics India Magazine, places AMD alongside NVIDIA and Broadcom. These three chipmakers have expressed a desire to buy up to $5 billion in Meta’s Class A offering. Their collective interest in funding a major AI company’s infrastructure shows hardware providers are seeking direct financial stakes in the AI build-out. For AMD, this move directly follows an $8.2 billion AI acquisition completed last month.

On September 30, AMD announced the acquisition of World Labs, a startup co-founded by AI pioneer Fei-Fei Li. This deal, first covered by Ars Technica, brought World Labs’ focus on “World Models” into AMD. World Models research involves AI systems that build internal simulations of their environments, allowing for more complex reasoning and planning. The $8.2 billion price tag for World Labs shows AMD investing heavily in advanced AI research and software capabilities. It moves the company beyond its traditional role as a hardware vendor, committing to own a piece of the AI software stack itself, not just the silicon.

The World Labs acquisition and the interest in Meta’s stock sale show AMD pursuing an aggressive AI strategy. The company is actively acquiring deep AI intellectual property and talent, spending $8.2 billion on World Labs. Simultaneously, it looks to embed itself financially within the AI infrastructure of major technology companies like Meta. AMD’s role is expanding to include both foundational AI model development and direct participation in the large-scale compute deployments that power these models. It invests in the intelligence itself and in the physical manifestation of that intelligence at scale.

This commitment to AI also comes as the broader AI compute market changes, now benefiting AMD’s core server CPU business. While GPUs have received significant attention for AI training, artificial intelligence agents have put the central processing unit (CPU) back in the spotlight. The complex orchestration, data pre-processing, and specific inference tasks performed by AI agents often require the general-purpose compute strengths of CPUs. Today, Analytics India Magazine also reported that AMD’s server CPU business is seeing renewed demand due to this AI agent boom. AMD’s EPYC server CPUs are a major contender in this market. The company gains an additional avenue to capitalize on the expanding AI sector beyond its GPU offerings. This market shift gives commercial rationale for AMD’s intensified AI investments, as it can gain from both the specialized GPU market and the revitalized CPU demand driven by AI.

The sequence of these events: an $8.2 billion AI software acquisition on September 30, followed by a potential multi-billion dollar investment in a major customer’s AI infrastructure in October, all while its core server CPU business gains new relevance from AI agents. This reveals AMD’s concerted effort to secure a dominant position in artificial intelligence. These are deliberate actions to embed AMD deeper into the AI value chain, from foundational models and research to the infrastructure that runs them, across both GPU and CPU domains. The company is positioning itself as a full participant in the AI arms race, not just a hardware supplier.

What company did AMD acquire in September 2026?

On September 30, 2026, AMD acquired World Labs, an AI startup co-founded by Fei-Fei Li.

How much did AMD pay for World Labs?

AMD paid $8.2 billion to acquire World Labs.

What is Meta doing to fund its AI infrastructure?

Meta is seeking to raise $95 billion through stock sales to fund its artificial intelligence infrastructure.

Which companies are interested in Meta’s stock offering?

NVIDIA, AMD, and Broadcom have collectively indicated an interest in purchasing up to $5 billion in Meta’s Class A offering.

Compiled by Launch91 Desk from the sources linked above. More about Launch91.