Radar / AI infrastructure / Microsoft Maia 300
Microsoft Maia 300: 3rd-gen Azure AI inference chip
Microsoft's third-generation in-house AI accelerator, announced Oct 12 at OCP Global Summit: TSMC N2P chiplets, 288GB HBM4 at 18TB/s, 950W, claiming 40% more tokens per megawatt; production deployments in 2027 with multi-gigawatt Azure scale-up.
Why it matters
Microsoft's biggest silicon bet yet: inference-first design optimized for long-context reasoning and multi-step agent workloads, aimed at making Maia the primary inference system across Copilot, Teams, and Azure services.
What you could build with it
A cloud-cost-conscious AI startup could track Maia 300's tokens-per-watt claims as a pricing signal: if Microsoft's inference costs drop in 2027, startups building on Azure could redesign agent workloads around cheaper long-context inference and pass the savings to customers.
Does it hold up?
Too early to judge: announced Oct 12 with production in 2027; no independent benchmarks, pricing, or hands-on tests exist. All performance figures are Microsoft's own claims.
Built with Microsoft Maia 300
- What is Maia 300? (kie.ai explainer)blog · Concise breakdown: 288GB HBM4, 18TB/s, 40% more tokens per megawatt, production 2027; notes Maia stays in-house, unlike Google TPUs.
- Microsoft announces Maia 300 (Unite.AI)article · Announcement coverage with Scott Guthrie's blog-post quotes on the inference-first architecture and multi-gigawatt scale-up plans.
- Microsoft unveils Maia 300 AI chip (Free Press Journal)article · Independent coverage of the Oct 12 OCP announcement: N2P chiplets, 288GB HBM4, 40% more tokens per megawatt.
- Microsoft Azure Maia 300 and Cobalt 200: AI spec guidearticle · Spec-focused analysis of Maia 300 within Microsoft's open-ecosystem infrastructure strategy.
Learn more
First spotted on article: source.
More AI infrastructure
DeepGEMM: DeepSeek's open GPU BLAS kernel libraryDeepSeek's DeepGEMM, a clean and efficient open-source BLAS kernel library for NVIDIA GPUs (up to 1550 TFLOPS on…infra · JEV 0.77OpenViking — self-evolving context database for AI agentsByteDance Volcengine's open-source (AGPL-3.0) context database that unifies agent memory, knowledge RAG, and skills…infra · JEV 0.76Modular open-sources MAX inference server, Mojo stdlib and accelerator kernelsThe core of Modular's unified AI deployment stack is now open: the MAX inference server (OpenAI-compatible endpoints)…infra · JEV 0.76OmniRouteFree MIT AI gateway: one endpoint, 290+ providers and 500+ models with quota-aware auto-fallback.infra · JEV 0.75Context Mode: context-window optimization for AI coding agentsMCP server for AI coding agents that sandboxes verbose tool output and persists session memory, cutting context use by…infra · JEV 0.74StrataOpen-source inference engine that runs the 125B-parameter Qwen3.8-Flash-Next MoE on a single consumer GPU with 12GB+…infra · JEV 0.74
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.