AMD Unveils MI450 AI Chips, Wins Microsoft and 2-Gigawatt Anthropic Deals
Updated
Updated · Yahoo Finance · Aug 25
AMD Unveils MI450 AI Chips, Wins Microsoft and 2-Gigawatt Anthropic Deals
1 articles · Updated · Yahoo Finance · Aug 25
Summary
AMD paired its new Instinct MI450 GPUs and 6th Gen EPYC Venice CPUs with two marquee customers: Microsoft will deploy Helios on Azure in second-half 2026, while Anthropic plans up to 2 gigawatts starting in first-half 2027.
The Anthropic tie-up could be worth tens of billions of dollars over time, and AMD also committed to invest up to $5 billion in the AI startup as it pushes harder against Nvidia in AI infrastructure.
Helios is AMD's new rack-scale AI system, combining GPUs, CPUs, networking and software into one platform for training and running large AI models.
The product push follows a strong Q2: revenue jumped 50% to $11.5 billion, data-center sales more than doubled to $6.7 billion, and AMD forecast second-half 2026 server CPU revenue growth of 80%.
Raymond James upgraded AMD to Strong Buy and raised its price target to $641 from $565, implying nearly 40% upside; the stock rose 5% in afternoon trading.
With a $5 billion Anthropic investment, is AMD buying its way to dethrone Nvidia in the AI supremacy race?
As gaming revenue plummets, can AMD's high-stakes pivot to agentic AI infrastructure truly justify a $641 price target?
How will the grid handle Anthropic's massive 2-gigawatt deployment of AMD's newest AI chips by 2027?
AMD’s Helios Revolution: $20 Billion in AI Deals, 20 GW of Deployments, and the Battle for Data Center Dominance
Overview
In August 2026, AMD made a bold move against Nvidia’s AI dominance by securing multi-billion dollar partnerships with Anthropic and Microsoft. These deals led hyperscalers to commit to gigawatt-scale deployments of AMD’s new Helios rack systems, pushing AMD’s disclosed pipeline to 20 gigawatts across top customers. Anthropic, fueled by a $5 billion AMD investment, agreed to deploy 2 gigawatts of Instinct GPUs and use its Claude model to help optimize AMD’s software stack. As Nvidia’s GPU supply remained a bottleneck, hyperscalers treated AMD as a crucial second source, gaining leverage in negotiations and accelerating hardware diversification across the AI industry.