Meta Strikes Up to $100 Billion AMD Deal for 6 Gigawatts of AI Chips with 160 Million Share Warrant Making Meta Major Shareholder

Meta Platforms will deploy 6 gigawatts worth of data center gear based on processors from Advanced Micro Devices in a blockbuster deal worth potentially up to $100 billion, marking a major win for the chipmaker's attempts to challenge Nvidia's dominance in AI infrastructure.

The five-year agreement, announced Tuesday, includes Meta purchasing AMD's MI540 series GPUs and latest-generation CPUs beginning in the second half of 2026. The series of transactions will be worth "double-digit billions" of dollars per gigawatt, according to AMD Chief Executive Officer Lisa Su, who declined to provide more specific valuation details.

As part of the arrangement, Meta will receive warrants to buy up to 160 million AMD shares—approximately 10% of the company—for $0.01 each, structured to vest alongside certain milestones. The full stock award is conditional on both project milestones and AMD's share price, which would need to hit $600 for Meta to receive its final tranche, according to The Wall Street Journal.

AMD's stock closed at $196.60 on Monday and rose as much as 8.6% after trading got underway Tuesday morning following the announcement. If fully vested, the warrant would transform Meta into a major AMD shareholder with significant influence over the chipmaker's strategic direction.

The AMD partnership represents the latest step in Meta's colossal spending spree as CEO Mark Zuckerberg makes AI the company's top priority. Meta has pledged to invest at least $600 billion in US data centers and AI infrastructure over the next several years, including a projected capital expenditure spend of $135 billion in 2026 alone.

The Facebook and Instagram parent company recently unveiled plans for a $10 billion gas-powered data center campus in Indiana designed for 1 gigawatt of compute capacity. The AMD deal would provide sufficient hardware to power six facilities of that scale.

Zuckerberg has committed to "aggressively front-load" computing capacity to maintain competitive positioning as the AI race intensifies. The company's AI investments span model development, consumer products like Meta AI assistant, business tools for advertisers, and infrastructure to support hundreds of millions of daily AI interactions across its platforms.

The AMD agreement comes just weeks after Meta struck a multiyear deal to expand its data centers with millions of Nvidia's latest CPUs and GPUs. The dual-vendor strategy reflects Meta's desire to diversify beyond sole reliance on Nvidia, which has dominated AI chip supply and commanded premium pricing.

CPUs are increasingly becoming a core pillar of the AI inference compute stack because they're efficient, easier to scale, and don't tie companies solely to Nvidia's ecosystem. AMD's MI540 series represents the company's latest effort to close the performance gap with Nvidia's H100 and H200 GPUs that currently power most large-scale AI training and deployment.

Meta is also working on its own in-house chips but has reportedly hit delays, making external partnerships even more critical to meeting near-term capacity needs.

For AMD, the Meta deal provides massive revenue visibility and validates the company's AI chip strategy. AMD's Data Center segment generated $5.38 billion in Q4 2025, up 39% year-over-year, with full-year 2025 revenue reaching $34.64 billion, up 34%.

CEO Lisa Su noted when entering 2026 that compute demand continues significantly outpacing supply across the industry, creating tailwinds for AMD's datacenter products. The Meta partnership locks in a substantial portion of AMD's production capacity through 2031.

The warrant structure also aligns AMD's incentives with Meta's success—if AMD's chips underperform or delivery schedules slip, Meta's equity stake becomes less valuable. Conversely, strong execution accelerates warrant vesting and rewards AMD with a higher stock price.

The deal underscores the extraordinary capital requirements of the current AI buildout phase. When a single customer commits potentially $100 billion over five years for chips and servers, it illustrates both the scale of hyperscaler ambitions and the winner-take-most dynamics emerging in AI infrastructure.

Bridgewater Associates estimated this week that Alphabet, Amazon, Meta, and Microsoft will collectively invest about $650 billion in AI-related infrastructure in 2026, up sharply from $410 billion in 2025. Meta's AMD purchase represents a significant portion of that total spending.

The announcement also signals growing urgency around chip diversification as companies recognize concentration risk in Nvidia dependency. While Nvidia remains the performance leader, AMD's improving capabilities combined with supply chain resilience make dual-sourcing increasingly attractive for hyperscalers planning decade-long infrastructure investments.

For the broader market, Meta's willingness to commit $100 billion to AMD indicates confidence that AI workload growth will justify the massive capital deployment—and that current model capabilities represent just the beginning of a multi-year scaling curve requiring exponentially more compute.