AMD introduced its new AI rack system, Helios, at the company’s Advancing AI conference in San Francisco, positioning it as a direct challenge to rival Nvidia in the market for high-performance computing infrastructure. CEO Dr. Lisa Su described Helios as the industry’s highest-performing AI rack, built to train and run demanding frontier AI models at massive scale, with deployment set to begin later this year.
Helios has already secured major customers, including OpenAI, Meta, Oracle, Anthropic, and Microsoft. Microsoft CEO Satya Nadella confirmed plans to expand Azure infrastructure using the system, while AMD and Anthropic announced a strategic partnership to deploy up to two gigawatts of GPUs through Helios. Reports indicate the system outperforms Nvidia’s competing Vera Rubin racks on several benchmarks.
AMD also unveiled its Venice-X CPU, designed for data center workloads and expected to launch in 2027.
Su said the rise of agentic AI, which requires extensive multi-step reasoning and tool use, is driving a significant increase in compute demand. She projected the AI accelerator market will reach approximately $1.4 trillion by 2030, nearing the size of today’s entire semiconductor industry, with GPUs expected to continue dominating due to their programmability amid rapidly evolving AI workloads and algorithms.
