When Meta announced its expanded partnership with Nvidia last week, the tech world saw it as a straightforward bet on the dominant player in AI chips. But just days later, Mark Zuckerberg’s company revealed a second, equally significant deal—this time with Nvidia’s biggest rival, AMD. The move signals a dramatic shift in how tech giants are approaching the AI infrastructure arms race. “This is about making the right bets at the right time” — Lisa Su, AMD CEO A Gigawatt Bet on Competition Meta’s multiyear deal with AMD involves deploying up to 6 gigawatts of the company’s graphics processing units for artificial intelligence data centers, along with AI-optimized central processing units. Early shipments of MI450 GPUs in AMD’s Helios rack-scale servers will begin later this year. The agreement includes a performance-based warrant for Meta to acquire 160 million AMD shares, representing approximately 10% of the company. The first tranche vests when the first 1GW of Instinct GPUs are shipped, with additional tranches tied to purchase milestones up to 6GW. The Warrant Structure Lisa Su described the warrant structure as a “win-win” for shareholders, underpinning what she called a “very ambitious” plan and financial model. She views the agreement as one of the “most transformational deals” for the chipmaker as it expands its AI capabilities. The deal mirrors a similar agreement AMD struck with OpenAI in October, which also gave the AI lab warrants to acquire 160 million shares of AMD. That agreement was tied to deployment and stock price benchmarks, establishing a template for how AMD is competing with Nvidia. Meta’s capital expenditure commitment this year reaches up to 35 billion as the company tries to keep pace with megacap peers as well as OpenAI and Anthropic in the global AI race. Overall, Meta has plans for 30 data centers, including 26 in the U.S. “We’re early in the cycle of seeing what the ultimate payoff can be. We have to invest ahead of the curve and really point in the direction that is going to have the largest benefit.” — Lisa Su, AMD CEO Market Response and Competitive Landscape AMD stock climbed 7% on the news Tuesday, while Meta shares traded slightly lower and Nvidia shares were largely flat. The market reaction reflects investor recognition that AMD is becoming a viable second option for AI giants and hyperscalers. Chip analyst Ben Bajarin of Creative Strategies estimates the deal is worth tens of billions of dollars over at least four years. “Six gigawatts would take quite some time to deploy,” he noted. A key differentiator from Meta’s Nvidia pact is that the first deployment involves customized GPUs—something analysts say Nvidia isn’t currently offering. The Road Ahead AMD’s Helios represents the first large-scale competition to Nvidia’s Grace Blackwell systems, which have experienced soaring demand since their launch in 2024. Nvidia controls roughly 90% of the AI chip market with a .66 trillion valuation, while AMD is valued at 20 billion. Meta has long been a customer of both AMD and Nvidia, and also develops in-house processors. According to reports late last year, Meta had been in talks with Google about deploying the search company’s tensor processing units in Meta data centers in 2027. The dual-track approach—committing to both Nvidia and AMD—reflects a broader industry reality: even the largest tech companies are hedging their bets in a market where compute capacity has become the primary constraint on AI development. This article was reported by the ArtificialDaily editorial team. For more information, visit CNBC. Related posts: Fractal Analytics’ muted IPO debut signals persistent AI fears in Indi Fractal Analytics’ muted IPO debut signals persistent AI fears in Indi India’s AI Moment: Fractal’s Muted IPO and a $1.1B Government Bet EY Identifies 10 Critical Opportunities as Tech Enters ‘Hyper-Velocity AI Moment’ Post navigation The AI Automation Gap: Why 76% of Finance Leaders Are Investing But On AI Models Deployed Nuclear Weapons in 95% of War Game Scenarios