Microsoft launches Azure Maia 300 for cloud AI inferencing

Azure Maia 300 is Microsoft's largest custom chip for Azure Cloud.

A dimly lit server rack with subtle blue LEDs pulsing, a lone monitor displaying code and data flow diagrams
AI-generated illustration

Microsoft has launched the Azure Maia 300, its largest custom processor for deployment on Azure Cloud. This new chip will serve as the primary AI inferencing system for Microsoft’s own products and services. The company plans to expand its Maia family of chips across multi-gigawatt deployments.

The Azure Maia 300 follows the Maia 200, which is currently being deployed across the US, Europe, and Australia regions. According to TechPowerUp, the Maia 200 delivers 70% more tokens per Watt across Copilot and Teams than AI GPUs.

Key facts

  • Azure Maia 300 launched by Microsoft
  • Maia 200 deployed in US, Europe, Australia
  • Maia 200 offers 70% more tokens/Watt vs. AI GPUs

Sources

  • TechPowerUpMicrosoft Releases Maia 300 Compute Processor for Azure Cloud

More news