Microsoft has officially announced the latest Microsoft Maia 200 and it is an AI inference accelerator built on 3nm process with native FP8/FP4 tensor cores. It comes with a 750W SoC TDP envelope along with a redesigned memory system with 216GB HBM3e at 7 TB/s and 272MB of on-chip SRAM.
The newly announced Microsoft Maia 200 is joining with Microsoft’s portfolio of CPUs, GPUs, and custom accelerators. It is actually offering more options in the hands of customers to run advanced AI workloads faster and more cost-effectively on Azure.
Microsoft will be offering multiple models with the new Maia 200 AI inference accelerator. It includes the latest GPT-5.2 models from OpenAI, bringing a performance per dollar advantage to Microsoft Foundry and Microsoft 365 Copilot. It is also revealed that each Microsoft Maia 200 chip contains over 140 billion transistors and is tailored for large-scale AI workloads.
The Microsoft Maia 200 also introduces a novel, two-tier scale-up network design built on standard Ethernet. Each accelerator exposes: 2.8 TB/s of bidirectional, dedicated scaleup bandwidth; predictable, high-performance collective operations across clusters of up to 6,144 accelerators.
Microsoft is inviting developers, AI startups, and academics to begin exploring early model and workload optimisation with the new Maia 200 software development kit (SDK).
Read interesting news, reviews as well as tips & tricks on TechnoBugg website, and stay updated with the latest happenings of the tech world on the go with Technobugg App. Also follow on Google News and join our Telegram channel as well as WhatsApp Channel for the latest updates.
