AI Semiconductors
Microsoft Maia
Microsoft's custom AI accelerator, with the second-generation Maia 200 launched in January 2026 delivering over 10 PFLOPS of FP4 compute on a TSMC 3nm process, designed as a first-class citizen within Azure with custom firmware and compiler toolchains optimized for OpenAI workloads.
Sources
Every claim traced to a primary source — evidence, recent activity, and insider filing behavior, all in one place.
Dependency Map
What Microsoft Maia depends on below, and who depends on Microsoft Maia above — click any node to make it the new center, 2 levels deep.
Connection type
Click any node to make it the new center. Scroll to zoom, drag to pan.
The Story So Far (last 6 months)
Auto-generated from this entity's dated milestones, relationship updates, and sourced evidence — not AI-written, just sorted.
Microsoft Maia's Timeline
A sourced, dated history of Microsoft Maia's key moments — founding to present.
Nov 2023 · Maia 100 unveiled
Microsoft unveiled its first-generation Maia AI accelerator as part of its custom-silicon strategy to reduce third-party chip dependency.
Jan 2026 · Maia 200 launched
Microsoft launched the second-generation Maia 200, delivering over 10 PFLOPS of FP4 compute on a TSMC 3nm process, designed as a first-class citizen within Azure.