Last week, Nvidia (NASDAQ:NVDA | NVDA Price Prediction) and Meta Platforms (NASDAQ:META) expanded their partnership, with Meta committing to millions of Nvidia GPUs for its AI data centers. The news sent Advanced Micro Devices (NASDAQ:AMD) stock tumbling about 3%, as some analysts speculated it could signal the end of AMD’s viability as a credible challenger to Nvidia’s dominance in the high-stakes AI accelerator market.
This morning, though, AMD delivered a resounding response by announcing its own blockbuster, multi-year agreement with Meta Platforms. The deal not only counters the narrative of Nvidia’s unchallenged lead but also highlights Meta’s deliberate strategy to diversify its AI compute suppliers amid exploding demand for training and inference capabilities.
Landmark Scale of the AI Partnership
The comprehensive, multi-generation pact calls for Meta to deploy up to 6 gigawatts (GW) of AMD Instinct GPUs across its data centers — an enormous scale that ranks among the largest AI chip supply agreements ever disclosed. Shipments for the initial 1-GW phase are slated to begin in the second half of 2026. This builds on the companies’ existing collaboration, which already incorporates prior AMD Instinct MI300 and MI350 series GPUs, along with EPYC CPUs.
At the heart of the agreement are custom AMD Instinct GPUs based on the upcoming MI450 architecture, specifically optimized for Meta’s demanding AI workloads, including large-scale model training and efficient inference. The partnership also encompasses sixth-generation AMD EPYC “Venice” CPUs, the open-source ROCm software stack, and the jointly developed AMD Helios rack-scale architecture through the Open Compute Project. AMD and Meta have committed to aligning future roadmaps across silicon, systems, and software, ensuring tailored, high-performance, and energy-efficient solutions for Meta’s gigawatt-scale AI infrastructure needs.
Valuation estimates for the deal vary widely but underscore its magnitude. Reuters reports suggest AMD could sell up to $60 billion worth of AI chips to Meta over five years. Other sources, including Bloomberg, The Wall Street Journal, and market analyses, place the potential value in the tens of billions, with some projections exceeding $100 billion. AMD CEO Lisa Su has referenced double-digit billions per gigawatt in commentary on comparable large-scale agreements, reinforcing the transformative revenue impact.
AMD’s leadership emphasized that the partnership will fuel substantial multi-year revenue growth and prove accretive to non-GAAP earnings per share. CFO Jean Hu described it as a key milestone in executing the company’s long-term financial ambitions.
This positions AMD firmly as a cornerstone supplier in Meta’s multi-vendor AI strategy, which already includes its Nvidia partnership and in-house MTIA accelerators. The sheer scale reflects intensifying demand for diverse, cost-effective, and power-efficient alternatives in the AI chip market, where Nvidia’s near-monopoly has driven hyperscalers to seek competitive options to mitigate supply risks, pricing power, and energy consumption challenges.
Meta’s Potential Equity Stake
To further cement the relationship, AMD issued Meta a performance-based warrant for up to 160 million shares of common stock — equivalent to roughly 10% of shares outstanding. Vesting occurs in tranches linked to Instinct GPU shipment milestones, starting with the first 1-GW phase, and includes additional conditions such as AMD stock price targets and Meta achieving specified technical and commercial goals.
However, Meta holds no shares under this agreement currently. This structure mirrors AMD’s prior arrangement with OpenAI and effectively aligns incentives for sustained collaboration and mutual success.
The announcement sparked a powerful rally in AMD shares, surging more than 10% in premarket trading. Investors see the deal as major validation of AMD’s technology roadmap and its growing traction with top-tier hyperscalers.
Key Takeaway
Last week, I wrote that the Nvidia-Meta expansion was overhyped as a potential knockout blow to AMD, yet proved a classic buying opportunity for savvy investors. Today’s landmark agreement with Meta emphatically reinforces that perspective.
By securing one of the industry’s most ambitious AI infrastructure commitments, AMD not only reasserts itself as Nvidia’s strongest challenger but also demonstrates the accelerating shift toward a more competitive, diversified AI accelerator ecosystem.
In a market where no single vendor can fully meet hyperscaler needs, this deal signals AMD’s momentum and long-term potential in the AI era.
Top Gaining Stocks
Top Losing Stocks