Microsoft unveils new AI chip to reduce dependence on nvidia

Microsoft has introduced its third-generation Maia AI chip, designed to serve as the primary AI inference system for its products and services. The Maia 300 is intended to enhance the cost and efficiency of running AI workloads in the cloud, while also being more capable of handling advanced reasoning models and agentic tasks. The chip is set to begin production rollout in 2027. Scott Guthrie, executive vice president of Microsoft's cloud group, highlighted that Maia 300 is expected to deliver 40 percent more tokens per megawatt compared to leading GPU-based systems. The Verge reports the full story. The Maia 300 represents Microsoft's continued investment in in-house AI infrastructure, aiming to reduce its reliance on Nvidia's GPUs. This development underscores the growing competition in the AI chip market, with major tech companies seeking to enhance their capabilities through proprietary solutions. The chip's efficiency improvements are expected to have a significant impact on cloud computing operations, particularly as AI workloads continue to expand.