AMD launches Helios AI rack-scale system to challenge Nvidia
Original: AMD takes on Nvidia with its Helios AI rack-scale system
Why This Matters
AMD's entry into rack-scale AI systems intensifies competition in a market projected to reach $1.4 trillion by 2030.
AMD unveiled its Helios AI rack-scale system at its sold-out Advancing AI conference in San Francisco on July 23. CEO Lisa Su called it the industry's 'highest-performance AI rack,' with customers including Microsoft, OpenAI, Meta, Oracle, and Anthropic. Shipments are planned for later in 2026.
AMD Chair and CEO Dr. Lisa Su unveiled the Helios rack-scale AI system at the company's sold-out Advancing AI conference in San Francisco on Thursday, positioning it as a direct challenger to Nvidia's Vera Rubin and Grace Blackwell systems. Su described Helios as 'built to train and run the most demanding frontier models in the world at massive scale,' with planned deployment at gigawatt-scale by leading AI companies. Performance benchmarks reported by The Register suggest Helios outperforms Vera Rubin on several metrics. Key customers include OpenAI, Meta, Oracle, Anthropic, and Microsoft. Microsoft CEO Satya Nadella announced Monday that Azure infrastructure would expand using Helios, while Anthropic and AMD confirmed a strategic partnership Wednesday to deploy up to two gigawatts of GPUs via the system. AMD also introduced the Venice-X CPU, targeting data center and high-compute workloads, with a 2027 launch expected. Su projected the AI accelerator market will reach approximately $1.4 trillion by 2030—approaching the size of today's entire semiconductor market—driven largely by the rise of agentic AI, which she said demands significantly more GPU compute per task.