SignalCommodities
Microsoft is scaling Azure Maia to multi-gigawatt deployments from 2027 and making it the primary inference system for its products. Maia 200 is already in production on Copilot; Maia 300 targets 40%
Microsoft expanding Azure Maia to multi-gigawatt deployments signals major AI infrastructure scale-up.
Context
What happened: Microsoft said Azure Maia 200 is now in production and that it will scale the Maia family to multi-gigawatt deployments starting in 2027, with Maia 300 coming that year as its primary inference system.
Key numbers: Maia 300 expected to deliver 40% more tokens per megawatt than leading GPU-based systems; up to 7x rack-scale token throughput of Maia 200 in a 96-accelerator
Why it matters: Microsoft positions Maia as the primary inference system for its products and services, with Maia 200 already serving Copilot and Maia 300 designed for long-context, agentic workloads.
What changed
Microsoft said Azure Maia 200 is now in production and that it will scale the Maia family to multi-gigawatt deployments starting in 2027, with Maia 300 coming that year as its primary inference system.
Why it matters
Microsoft positions Maia as the primary inference system for its products and services, with Maia 200 already serving Copilot and Maia 300 designed for long-context, agentic workloads.