Daily Tech Wizard Newsletter

Stories / / By Daily Tech Wizard

Microsoft's Maia 300 chip rack: up to 7x Maia 200 output

Microsoft says its next in-house AI chip rack makes up to 7x more output.

Microsoft plans to run its own Maia AI chips at multi-gigawatt scale starting in 2027. CEO Satya Nadella calls it "our biggest bet yet on our own silicon."

Executive Scott Guthrie writes that Maia will become the company's main inference system across its products.

The current chip, Maia 200, already runs Microsoft Copilot in the US, Europe, and Australia. GitHub Copilot and Foundry follow by the end of the year. It handles OpenAI, MAI, and open-source models. Microsoft says it pushes up to 1.7 times more tokens per watt than leading GPUs in Teams.

Its successor, Maia 300, ships in 2027 and targets long inputs, fast chats, and agentic tasks. Each chip packs eight HBM4 memory stacks totaling 288 GB, with 18 TB/s of bandwidth, and is rated at 950 watts.

Microsoft claims 40 percent more tokens per megawatt than leading GPU systems. A 96-chip rack should produce up to seven times the tokens of a Maia 200 rack.

A technical report covers the details, with more coming in November at Ignite 2026.

Do you think more companies will build their own AI chips?

Video: Microsoft Azure on YouTube