Industry change
Microsoft to scale its Azure Maia AI silicon to multi-gigawatt deployments
- Published
- Source
- Microsoft Official Blog
Summary
Microsoft announced it will expand its homegrown Azure Maia accelerator family to multi-gigawatt deployments starting in 2027, making Maia the primary inference system for its products and services. The second-generation Maia 200 already serves production workloads including Microsoft Copilot, and the third-generation Maia 300 is slated for production in 2027. By the company's figures, Maia 300 is expected to deliver 40% more tokens per megawatt than leading GPU-based systems.
Why it matters for our work
Inference cost and power efficiency are core constraints on sustainable AI services. A major cloud improving them with its own silicon directly affects the unit economics and scaling speed of agent-based services.
Translated from the Korean original. Summaries may be translated and edited. Commentary reflects our perspective; forecasts remain the source’s views.