Microsoft to deploy Maya AI chips at scale in 2027
Microsoft plans to use its proprietary Maya chips for the majority of its AI inference workloads starting in 2027. The move aims to reduce reliance on external graphics processors while improving energy efficiency for products like Copilot.
Microsoft intends to deploy its own AI chips, the Maya series, at scale beginning in 2027 to handle the majority of inference workloads for AI models across its products. The current generation, Maya 200, is already operational in the US, Europe, and Australia, where it powers the Copilot assistant. Microsoft states that Maya 200 is significantly more energy-efficient than leading graphics chips, generating up to 1.7 times as much text per watt. The successor, Maya 300, is scheduled for release next year and is designed for long inputs, fast chats, and AI agents. The company claims the new chip will deliver up to seven times the performance of the previous generation within a single server rack. Further technical details regarding the deployment timeline and specifications are expected in November.