source: techcrunch ai: amd takes on nvidia with its helios ai rack scale system
level: business
amd introduced its helios ai rack-scale system at the advancing ai conference, positioning it as a direct competitor to nvidia's vera rubin and grace blackwell platforms. the system combines many processors into a single high-powered unit for data centers, targeting the largest ai labs. ceo lisa su called it the industry's highest performance ai rack, built to train and run frontier models at massive scale. helios will ship later this year with deployments at gigawatt-scale.
key customers include openai, meta, oracle, anthropic, and microsoft. microsoft plans to expand azure infrastructure with helios, while anthropic signed a deal for up to two gigawatts of gpus. amd also announced the venice-x cpu for data centers, launching in 2027. su projected the ai accelerator market will reach $1.4 trillion by 2030, driven by agentic ai workloads that require many gpus for multi-step reasoning and tool use.
performance metrics suggest helios beats nvidia's vera rubin on several counts, giving amd a real chance in a market nvidia has historically dominated. the push comes as ai compute demand shifts toward programmable gpus, which su says will make up most of the accelerator market. amd's growing customer list and strategic partnerships signal a serious effort to capture share in the expanding ai infrastructure space.
why it matters: more competition in ai hardware could lower costs and increase options for data science teams building and deploying large models.
source: techcrunch ai: amd takes on nvidia with its helios ai rack scale system