Microsoft Introduces Maia 200, Its Most Powerful AI Chip Yet

Microsoft is betting on a deeper kind of AI control… one most users never see.

On Monday, the tech giant introduced Maia 200, its newest custom-built AI workhorse. Designed specifically to handle the heavy lifting of running massive AI models, this chip serves as a direct challenge to the hardware crowns held by Amazon and Google.

The Maia 200 is built on a cutting-edge 3-nanometer process and packed with over 140 billion transistors. According to the Microsoft announcement, this translates into a massive jump in raw power, delivering over 10 petaFLOPS of 4-bit precision performance.

But for Microsoft, it’s not just about speed; it’s about the bottom line. The company claims this new silicon is its most efficient yet, offering 30% better performance-per-dollar than the hardware it currently uses. 

“Maia 200 is also the most efficient inference system Microsoft has ever deployed, with 30% better performance per dollar than the latest generation hardware in our fleet today,” said Scott Guthrie, Microsoft’s Executive Vice President of Cloud + AI.

Taking on the neighbors

Microsoft didn’t shy away from naming names. In its announcement, the company positioned the Maia 200 as the top dog among “hyperscaler” chips. 

“This makes Maia 200 the most performant, first-party silicon from any hyperscaler, with three times the FP4 performance of the third generation Amazon Trainium, and FP8 performance above Google’s seventh generation TPU,” according to the Microsoft blog.

While Amazon’s Trainium and Google’s TPUs have been around for years, Microsoft, which only entered the custom chip race in late 2023, is moving at breakneck speed to catch up and overtake them.

Already hard at work powering Copilot, GPT-5.2, and Microsoft’s own AI work

Maia 200 is not just a lab project. Microsoft says the chip is already running in its US Central data center region near Des Moines, Iowa, with deployment planned next for the US West 3 region near Phoenix, Arizona.

The accelerator will power Microsoft 365 Copilot and Azure workloads, including the latest GPT-5.2 models from OpenAI. Microsoft’s Superintelligence team will also use Maia 200 for synthetic data generation and reinforcement learning to improve future in-house models.

Developers get early access

In a sign of confidence, alongside the hardware, Microsoft is rolling out a preview of the Maia software development kit (SDK). The SDK includes PyTorch support, a Triton compiler, optimized kernels, a low-level programming language, and tools like a simulator and cost calculator.

Microsoft says the goal is to let developers, startups, and researchers optimize models early and move workloads smoothly across different AI accelerators in Azure.

Also read: AI trends to watch in 2026 include the push for AI infrastructure that gets smarter, not just bigger.

The post Microsoft Introduces Maia 200, Its Most Powerful AI Chip Yet appeared first on eWEEK.