Microsoft rolls out Maia 200 AI chip, positions it as alternative to Nvidia amid cloud rivalry
Microsoft has announced its second-generation Maia 200 AI chip, built on TSMC’s 3-nanometre process, with early deployments expected for internal frontier-model work and broader use across Copilot and Azure AI offerings—an escalation in the race against Nvidia and cloud competitors.
A new in-house AI chip from Microsoft
Microsoft has rolled out its second-generation artificial intelligence chip, Maia 200, as part of its push to reduce dependence on third-party AI processors and strengthen its position in the increasingly competitive cloud AI market. The company is pitching Maia 200 as a credible alternative to leading AI chips from Nvidia, while also signalling intent to compete more aggressively with cloud rivals such as Amazon and Google.
The Maia 200 arrives after the earlier Maia 100, which Microsoft had announced but did not make available for public cloud clients to rent. According to the report, Maia 200 is built using Taiwan Semiconductor Manufacturing Co’s 3-nanometre manufacturing process, pointing to a focus on efficiency and performance for large-scale AI workloads.
Where Maia 200 will be used first
Some of the first units are expected to be deployed within Microsoft’s Superintelligence team led by Microsoft AI CEO Mustafa Suleyman. The chip is also expected to support Microsoft’s Copilot assistant for businesses and power AI models that Microsoft offers to cloud customers, including the latest OpenAI models available through its cloud ecosystem.
For enterprise users in India and globally, the move matters because it could eventually influence pricing, capacity availability and performance consistency for AI services on Microsoft’s cloud. The broader industry implication is a sharper battle over the hardware stack itself, as hyperscalers try to control costs and differentiate their AI platforms beyond software alone.
- Maia 200 positioned as an alternative to Nvidia AI processors
- Built on TSMC 3-nm process
- Early use: Microsoft’s Superintelligence team
- Expected to power Copilot and cloud AI model offerings