Chips & Compute5 min reading time

Microsoft Commits Maia to Multi-Gigawatt Inference, and Its Efficiency Metric Draws Questions

InfoQ (AI, ML & Data)
Read the full article
Microsoft plans to expand its Azure Maia AI inference system to multi-gigawatt scale in 2027, making it central to Microsoft products. Maia 300 focuses on inference with enhanced memory bandwidth and a custom AI network fabric for efficient data movement.

More on this story


More in Chips & Compute

TSMC’s Upbeat Outlook Fails to Quash AI Buildout Concerns

Covered by 2 sources

Taiwan to Curb AI Chip Exports to China as Part of US Trade Deal

Bloomberg

Taiwan to Fund $4 Billion Texas Complex for AI, Nvidia Partners

Bloomberg