Microsoft plans to scale its in-house Azure Maia AI chip family to multi-gigawatt deployments from 2027, making Maia the main inference system for its products and services.
Microsoft plans to scale its in-house Azure Maia AI chip family to multi-gigawatt deployments from 2027, making Maia the main inference system for its products and services.
Maia 200 already powers services such as Copilot, running OpenAI models and Microsoft's own MAI models, while Maia 300, due in production in 2027, is expected to deliver 40% more tokens per megawatt than leading GPU-based systems. Built on TSMC's 2-nanometre-class process, Maia 300 is designed for long-context reasoning and multi-step agentic tasks.