← October 13, 2026 briefing · added in the 12:00 KST update

P1 Infrastructure Announced Efficiency & infraFollow-up12:00 updateMicrosoftNVIDIA

Microsoft to deploy its in-house Maia AI chips at multi-GW scale from next year; Maia 300 volume deployment in 2027 (reported)

Announced October 12, 2026

Verified (fact)

On October 12 (local time), CEO Satya Nadella said on X and the official blog that Microsoft is making its largest-ever investment in its in-house AI chip, Azure Maia. What is new beyond yesterday's Maia 300 announcement is the deployment plan. Starting next year, the Maia family will be deployed at a scale of multiple GW. Maia 200 is currently used in production services such as Microsoft Copilot. The next-generation Maia 300 will begin volume deployment in 2027 and is to become the main inference system for the company's products and services.

Needs fact-check (claim)

Microsoft claims Maia 300 delivers 40% more tokens per watt than leading GPU-based systems (the comparison systems and measurement conditions are not given in the reports).

Why it matters (interpretation)

A hyperscaler pledging to deploy its own chips by the gigawatt signals that the push to lower inference costs and reduce dependence on NVIDIA GPUs is moving into large-scale deployment.

Confidence medium · official source pending

Sources

Story thread

  1. 2026-10-12 Microsoft announces Maia 300, its third-generation in-house AI chip, as the 'primary inference system' for its services (report)
  2. 2026-10-13 Microsoft to deploy its in-house Maia AI chips at multi-GW scale from next year; Maia 300 volume deployment in 2027 (reported)

More about Microsoft