Slashdot has reported:

Microsoft has unveiled its Azure Maia 300 AI chip as it seeks to reduce reliance on processors from Nvidia and AMD. The announcement came Monday at the OCP Global Summit 2026 in San Jose. Microsoft said the Maia 300 will join its Maia 100 and Maia 200 chips in multigigawatt deployments beginning next year, making the Maia line the “primary AI inference system for Microsoft products and services.”

The company said Maia 300 delivers 40 per cent more tokens per megawatt than “leading GPU-based systems.” It also said a rack containing 96 chips provides seven times the rack-scale token throughput of a Maia 200 rack.

The Maia 200 currently powers Microsoft production workloads, including Microsoft 365 Copilot, along with OpenAI models, Microsoft’s MAI models and open-source models. Microsoft said Maia 200 provides 1.7 times more tokens per watt for AI capabilities in Teams, Copilot and Cowork software compared with other GPU-based systems.

The Maia 300 is the latest addition to Microsoft’s effort to build its own AI processors. The company said the chips are intended to support its products and services as it expands deployments starting next year.