Microsoft’s custom AI chip program is back on track, with the Maia 200 inference chip, built on TSMC’s 3nm process, already deploying in Microsoft data centers in Iowa and Arizona. The Maia 200 ships with 216GB of HBM3e memory and native support for FP8 and FP4 precision formats. Microsoft claims the chip delivers three times the FP4 performance of Amazon’s Trainium3, exceeds Google’s TPUv7 in FP8 performance, and offers roughly 30% better performance-per-dollar compared to the previous generation of Azure fleet silicon. Microsoft 365 Copilot and OpenAI’s models are running on Maia 200 hardware. The Maia program first surfaced in November 2023, but design changes and staffing challenges forced timeline shifts, pushing mass production to 2026. The Maia 200 was formally announced on January 26, 2026. The successor chip, Maia 300, is already in design, with a development timeline stretching into 2027. Discussions in mid-2026 indicated potential collaborations around external use of Maia 200 hardware, with Anthropic mentioned as a possible partner. Microsoft executives Scott Guthrie and Satya Nadella have emphasized how custom chips improve Azure AI’s cost structure. Microsoft’s push follows a pattern among major cloud providers, including Amazon, Google, and Meta, to reduce dependence on Nvidia, which commands the vast majority of the AI accelerator