Microsoft plans multi-gigawatt Maia chip deployments starting 2027
Article timeline
The articles in this story. Times are ET.
Rohan Paul@rohanpaul_aiXMicrosoft plans Maia 300 as its primary inference chip from 2027AIMicrosoft says it will scale its Maia chip family to multi-gigawatt deployments and make it the primary inference system for its own products. The company claims Maia 300 delivers 40% more tokens per watt than leading GPU-based systems. Reported talks with TSMC for more than 300,000 units remain under negotiation and are not a signed order.
Satya Nadella@satyanadellaOfficialMicrosoft plans multi-gigawatt Azure Maia system deploymentsAIMicrosoft says it will expand its Azure Maia system to multi-gigawatt deployments. The main post contains only a link to the Microsoft blog, so no further figures or timelines are available from this source.
- Satya Nadella@satyanadellaOfficialMicrosoft scales Maia silicon to multi-gigawatt inference and unveils Maia 300
AIMicrosoft says it will scale its Maia chip family to multi-gigawatt deployments starting next calendar year, making it the primary inference system for its products and services. The company is introducing Maia 300, which it says delivers 40% more tokens per watt than leading GPU-based systems. Maia 200 is already deployed in the US, Europe, and Australia, delivering up to 1.7x more tokens per watt across Teams and Copilot experiences.
