Meta could trade AI chips for shares in AMD under new partnership
Meta’s deal with AMD could give it a 10% stake as it secures massive AI chip capacity and deepens its push beyond NVIDIA.
Meta has agreed to a major artificial intelligence infrastructure deal with AMD that could see the social media giant acquire a significant stake in the chipmaker, highlighting growing ties between AI developers and semiconductor companies.
Table Of Content
The companies said Meta will buy up to 6 gigawatts of AMD’s Instinct graphics processing units, with the agreement structured so that AMD could issue Meta up to 160 million shares of common stock. If all milestones are met, Meta could end up owning up to 10% of AMD.
The deal underscores the escalating race to secure advanced AI hardware. It reflects a broader shift by major technology firms to diversify away from NVIDIA, which currently dominates the AI accelerator market.
A complex deal blending chips and equity
Under the agreement, Meta plans to deploy AMD’s Instinct GPUs based on the MI450 architecture, which have been optimised for Meta’s AI workloads. The first one gigawatt of capacity is expected to be deployed in the second half of 2026, with the remaining capacity rolled out as Meta scales its AI infrastructure.
The equity component of the deal is tied to both technical and commercial milestones. The first tranche of AMD shares would vest once the initial gigawatt of shipments is delivered, with additional tranches released as Meta expands its deployments to the full six gigawatts. Vesting is also linked to AMD achieving certain stock price thresholds.
Meta and AMD are also expanding their existing partnership around server processors. Meta plans to deploy millions of AMD EPYC central processing units. She will become a launch customer for AMD’s sixth-generation EPYC chips, signalling deeper integration of AMD technology across Meta’s data centres.
The structure of the agreement resembles a deal AMD struck with OpenAI last year, in which the chipmaker offered equity in exchange for large-scale GPU purchases. Such arrangements reflect the enormous capital requirements of AI infrastructure, where demand for cutting-edge chips has surged far beyond supply.
Growing concerns over circular AI partnerships
Analysts have compared these arrangements to circular transactions that bind AI companies and chipmakers in tightly coupled financial and technological relationships. By exchanging chips for equity, companies effectively become both customers and shareholders, blurring traditional commercial boundaries.
Critics warn that this structure could amplify risks if expectations around AI growth fail to materialise. If demand for AI services or infrastructure slows, both sides could face significant financial exposure, especially given the high valuations attached to AI-focused firms and semiconductor companies.
At the same time, such deals illustrate the strategic importance of securing long-term access to computing power. AI models are becoming larger and more resource-intensive, pushing technology companies to lock in supply agreements years in advance. For chipmakers, these agreements provide predictable revenue streams and strengthen ties with some of their largest customers.
Meta seeks to diversify beyond NVIDIA
The agreement highlights Meta’s efforts to reduce its reliance on NVIDIA, whose GPUs currently dominate AI workloads across the industry. AMD has emerged as a key alternative supplier, and the deal could help the company expand its presence in data centre AI markets.
Meta said the partnership would help it build a more flexible infrastructure stack. “By diversifying our partnerships and technology stack, we’re building a more resilient and flexible infrastructure,” the company said in its news release.
The move also reflects a broader industry trend, with major cloud providers and AI developers increasingly exploring alternatives to NVIDIA’s hardware. As competition intensifies, companies are seeking to secure favourable pricing, reduce supply chain risks, and gain greater control over their AI roadmaps.
For AMD, the deal represents a significant opportunity to scale its AI business and challenge NVIDIA’s market leadership. However, it also ties the company more closely to the fortunes of its largest customers, potentially increasing volatility if AI spending fluctuates.
Overall, the agreement marks another milestone in the rapidly evolving AI ecosystem, where strategic partnerships, equity swaps and long-term supply commitments are becoming central to the battle for technological leadership.
