- AMD Surges 14% on Meta Contract: Shares of Advanced Micro Devices skyrocketed in U.S. pre-market trading following a massive five-year agreement to supply 6 gigawatts of AI infrastructure to Meta Platforms.
- By securing a "Tier-1" hyperscale supplier status with Meta, AMD has successfully positioned itself as the primary alternative to Nvidia in the rapidly expanding AI accelerator market.
The AI infrastructure race hit a new gear on Tuesday, February 24, 2026, as Advanced Micro Devices (AMD) shares surged more than 14% in U.S. pre-market trading. The rally comes on the heels of a massive five-year partnership with Meta Platforms (META), a deal that officially elevates AMD into the elite “Tier-1” group of AI hyperscale suppliers.
As reported by Bloomberg, “Meta Platforms Inc. will deploy 6 gigawatts’ worth of data center gear based on processors from Advanced Micro Devices Inc., a blockbuster deal that marks a win for the chipmaker’s attempts to catch up with Nvidia Corp.”
Meta’s $100 Billion AI Infrastructure Deal Signals a New Capex Supercycle
The scale of the agreement is virtually unprecedented in the current capex cycle. Starting in the second half of 2026, Meta will begin deploying AMD-based AI infrastructure totaling 6 gigawatts (GW) of capacity.
According to AMD CEO Lisa Su, the value of these transactions is expected to reach “double-digit billions” of dollars per gigawatt. This implies a total program value that could exceed $100 billion over the five-year deployment horizon. For a company like AMD, which reported roughly $34.6 billion in revenue last year, this contract represents a monumental shift in its long-term revenue trajectory.
Meta Could Become a Major AMD Shareholder Through AI Chip Partnership
The partnership goes beyond a standard hardware purchase. The deal structure includes equity warrants for Meta to purchase 160 million AMD shares. If these warrants are exercised as project milestones are met, Meta could potentially build a 10% stake in AMD. This “co-design” model ensures that Meta gains direct influence over future chip roadmaps while optimizing hardware for its specific inference needs, the phase where trained AI models are deployed into production.
AMD Technical Chart Analysis
AMD is currently trading at $209.61, reflecting a sharp weekly bounce.
Key Resistance Levels
- Immediate Resistance $213.00 – $220.00: The stock is currently testing the weekly high of $213.43.
- Major Resistance $260.00 – $270.00
- Immediate Support $194.00 – $198.00: This is the most critical floor for the current uptrend.
- Primary Support $180.00: Should the $190 level fail, the next structural support is found near the $180 mark.
Meta’s AI Supercycle Strategy: Why Massive Capex Still Makes Sense
Despite ongoing debates regarding an AI “capex bubble,” Mark Zuckerberg’s broader strategy remains clear: diversification and massive scale. Meta’s total AI infrastructure spending is expected to hit $135 billion this year alone. By adding AMD to its mix alongside Nvidia and its own in-house silicon, Meta is ensuring it has the energy-efficient capacity required to lead the next decade of AI monetization.
The deal is officially valued at $60 billion to over $100 billion over a five-year period. AMD CEO Lisa Su confirmed the transaction series is worth “double-digit billions” per gigawatt, and with a total commitment of 6 gigawatts, the total contract value could potentially exceed $120 billion depending on the hardware mix of custom GPUs and CPUs.
To secure the deal, AMD issued Meta performance-based warrants for 160 million shares at an exercise price of just $0.01.
These shares only vest as Meta hits specific shipment milestones (1 GW scaling to 6 GW) and as AMD’s stock price hits rising targets up to $600.
While this could eventually give Meta a 10% stake in AMD, investors view it as “good dilution” because the warrants are tied directly to massive revenue realization and high stock price performance.
No. Meta is pursuing a “Multi-Sourcing” strategy. Meta recently reaffirmed its partnership with Nvidia to buy millions of Blackwell and Rubin chips.
AMD’s Role: AMD’s custom MI450 hardware is being optimized specifically for Inference (running AI models), while Nvidia remains the leader for the more intensive Training phase.
Meta is using AMD, Nvidia, and its own in-house silicon to avoid being dependent on a single vendor.