Meta is reportedly on track to start producing the latest versions of its AI-specific chips in September. The move is part of the company's effort to reduce spending on GPUs from suppliers such as Nvidia.
Custom AI silicon has become a strategic priority for major AI companies because compute costs directly shape model training and inference economics. Owning more of the stack can improve margins and supply certainty.
Meta's chip push adds to a broader industry trend in which hyperscalers and AI labs are looking for alternatives to buying every unit of compute from the same GPU supply chain.