Microsoft is preparing its next-generation Maia 300 artificial intelligence chip as the company moves to significantly expand its own AI hardware production.
The Maia 300 could be introduced as early as September 2026, although the exact launch timing has not been officially confirmed. The new processor will continue Microsoft’s push to design more of the hardware used to power artificial intelligence services across its cloud infrastructure.
One of the biggest parts of Microsoft’s plan is production scale. The company is reportedly discussing manufacturing capacity with Taiwan Semiconductor Manufacturing Company, better known as TSMC, for more than 300,000 Maia 300 chips that could be delivered during 2027.
Microsoft’s longer-term ambitions could be much bigger. The company is said to be interested in securing enough manufacturing capacity for more than one million Maia 300 processors. However, reaching that level will depend on the availability of important chip components and manufacturing capacity.
Building its own AI processors gives Microsoft another way to handle the huge computing requirements created by modern AI systems. Microsoft currently uses large numbers of processors from companies such as Nvidia, whose GPUs have become some of the most important hardware for training and running advanced AI models.
Developing Maia chips could help Microsoft reduce some of that dependence while giving the company greater control over the cost, performance and design of its AI Infrastructure.
Microsoft is not expected to completely move away from outside chip suppliers. Instead, its custom processors can become another important part of the hardware available inside its data centers, alongside processors from Nvidia, AMD and other technology companies.
The Maia 300 follows the Maia 200, which Microsoft introduced earlier in 2026. Maia 200 is manufactured using TSMC’s advanced 3-nanometer technology and was designed with large amounts of fast on-chip memory to improve performance when AI systems are processing many requests.
With Maia 300, Microsoft appears ready to move its custom AI chip program toward a much larger production stage.
The company also wants its processors to become useful beyond Microsoft’s own internal services. Large AI developers and Azure customers could eventually become important users of Maia hardware as Microsoft works to make its cloud platform more competitive for demanding AI workloads.
That puts Microsoft in a growing custom-chip race with other major cloud providers. Google has spent years developing its Tensor Processing Units, while Amazon has expanded its Trainium processors for artificial intelligence workloads.
For Microsoft, producing hundreds of thousands of Maia chips could strengthen Azure by giving the company more control over the infrastructure behind AI services. It could also provide another option as demand for advanced AI computing hardware continues to grow.
Exact production numbers remain subject to manufacturing capacity, component availability and Microsoft’s final agreements with chip suppliers. Still, the reported plans show that Microsoft is preparing to make its own silicon a much larger part of its long-term AI and cloud strategy.
