AMD gains ground as AI agents shift compute needs away from Nvidia GPUs

As AI adoption has matured beyond training large models, new agentic applications like Meta's Muse and OpenAI's Dots are driving demand for different hardware. These agent-based systems rely primarily on virtual machines powered by central processors, where AMD and Intel have competitive advantages over Nvidia, which has dominated the GPU market for model training and inference.

This shift represents a significant departure from the early AI boom, which was heavily shaped by Nvidia's graphics processing units. The move toward agents running on CPUs could reshape the chip market and create new growth opportunities for processor makers who have traditionally competed on other fronts.