Microsoft unveils its third-gen in-house AI chip, the Maia 300, built on TSMC's N2P process node and aimed at serving as its "primary AI inference system" (Tom Warren/The Verge)
The new Maia 300 chip will start rolling out to production AI workloads in 2027.
Microsoft has announced its third-generation in-house AI chip, the Maia 300, which will begin rolling out to production AI workloads in 2027. The new chip is designed to be the primary AI inference system for Microsoft's products and services, with improved cost and efficiency for running AI workloads in the cloud. Maia 300 offers 40% more tokens per megawatt compared to leading GPU-based systems and up to 7 times the rack-scale token throughput in a 96-accelerator configuration, according to Scott Guthrie, executive vice president of Microsoft's cloud group.
This efficiency boost will allow Microsoft to rely more on its own AI chips in the future, reducing reliance on Nvidia and AMD. Maia 300 is a successor to the Maia 200, which is already in use for production AI workloads, including Microsoft Copilot, GitHub Copilot, Microsoft Foundry, and Microsoft's AI models. The chip is built using a new chiplet architecture on TSMC's N2P process node, with separate compute, memory controller, and I/O functions to enhance efficiency.
Maia 300 is essential for handling long-context reasoning, latency-sensitive conversational experiences, and sustained multi-step agentic tasks. Microsoft aims to catch up with industry leaders like Google DeepMind, OpenAI, and Anthropic by running its AI models on top of the Maia chip, a critical focus for their team.
Written by urgent.news from Techmeme's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
