Microsoft 自研芯片 Maia 300 将于 2027 年成为 Azure 主力推理系统
微软要把自家 Maia 芯片推到台前了，Maia 300 每兆瓦 tokens 比主流 GPU 高 40%，2027 年直接替代部分 Nvidia，价格战有得看。
Microsoft 计划从 2027 年起把 Azure Maia 作为自家产品和服务的主要推理系统。目前的 Maia 200 已支撑 Copilot 等生产负载，运行 OpenAI 模型、Microsoft 自研 MAI 模型和开源模型。Maia 300 将于 2027 年开始生产部署，Microsoft 预计其每兆瓦 tokens 产出比主流 GPU 系统高 40%。此举可能大幅降低 Microsoft 对 Nvidia GPU 的依赖。
Now its getting interesting! Microsoft is taking on Nvidia in AI inference with a multi-gigawatt expansion of its own chips.
Starting in 2027, Microsoft plans to make Azure Maia the primary inference system for its own products and services.
Maia 200 already serves production workloads including Microsoft Copilot, running OpenAI models, Microsoft's own MAI models and open-source models.
Maia 300 begins production deployments in 2027. Microsoft expects it to deliver 40% more tokens per megawatt than leading GPU-based systems.
Microsoft does not name Nvidia in the announcement, but the planned rollout could reduce its reliance on Nvidia GPUs for running its own AI services massively.
