AMD has officially introduced its next-generation MI400 AI accelerator chips, developed in partnership with OpenAI and its CEO, Sam Altman. According to CNBC, the MI400 chips are designed to rival Nvidia’s popular AI GPUs, which currently dominate the training and inference of large language models. CEO Lisa Su unveiled the MI400 series at an industry event, highlighting its improved power efficiency, higher computational throughput, and optimized software stack for running frontier AI models. This collaboration comes at a time when the AI chip market is under intense competitive pressure. Sam Altman, who heads OpenAI, has long advocated for more hardware innovation to keep up with rapidly advancing AI models. The partnership with AMD is a strategic move to diversify the AI compute supply chain and reduce dependence on Nvidia’s hardware. The MI400 series will directly target data centers and AI research labs working on models like GPT-5 and beyond. Early benchmark reports suggest that the MI400 performs significantly better than its predecessor, the MI300, with substantial gains in energy efficiency and total cost of ownership. OpenAI is expected to be among the first large-scale customers to deploy the MI400 chips in its future training clusters. Analysts believe this collaboration could signal a shift in the balance of power in the lucrative AI hardware industry. With growing geopolitical scrutiny of chip exports and supply chain resilience, AMD’s expanded AI portfolio may also help satisfy new global demands for diverse compute solutions. This announcement underscores AMD’s commitment to becoming a more formidable rival to Nvidia in AI and high-performance computing. For more insights, read the full story at CNBC.