Microsoft announces Maia 300 chip for multi-gigawatt AI inference deployments

Microsoft detailed the third-generation Maia 300 custom AI inference chip and stated it will scale Azure Maia systems to multi-gigawatt deployments beginning in 2027, making Maia the primary inference platform for Microsoft products and services. The company projects 40% more tokens per megawatt versus leading GPU systems and up to 7x higher rack-scale token throughput than Maia 200. This shifts Microsoft toward greater reliance on its own silicon for cost and efficiency in serving large-scale AI workloads.