Microsoft has announced Maia 300, its third-generation in-house AI chip, positioned as the “primary AI inference system” for the company’s products and services. According to Microsoft, the chip is built to improve the cost and efficiency of running AI workloads in the cloud and to better handle reasoning models and agentic tasks. Executive vice president Scott Guthrie said Microsoft expects Maia 300 to deliver 40 percent more tokens per megawatt compared with leading GPU-based systems. The chip will begin rolling out to production in 2027.
Why it matters
The move is framed as reducing Microsoft’s reliance on Nvidia’s GPUs for AI inference. A more efficient in-house chip could lower the cost and energy demands of running large AI workloads at cloud scale.
Who should care
Enterprises and developers running AI inference on Microsoft’s cloud, as well as those tracking the competitive dynamics between cloud providers and GPU suppliers, will want to follow the planned 2027 rollout.