Microsoft Could Unveil Maia 300 AI Chip as Early as September: Report

Microsoft is preparing to publicly unveil its next-generation Maia 300 AI chip this fall, potentially as soon as September, according to a Reuters report citing The Information, which cited people with direct knowledge of the company's plans.

A Bigger Bet on In-House Silicon

Microsoft first introduced its Maia AI chip line in November 2023 but has trailed rivals Alphabet and Amazon in scaling up its custom silicon efforts, as the company works to reduce its reliance on Nvidia's costly processors. Google began recognizing revenue from direct sales of its custom Tensor Processing Units in the quarter ended June, while Amazon has seen growing adoption of its own in-house Trainium chips — pressure that appears to be accelerating Microsoft's own timeline.

According to the report, Microsoft has been in talks with chipmaker TSMC to secure manufacturing capacity for more than 300,000 units of the Maia 300 for delivery in 2027. The company is reportedly aiming even higher over the longer term, targeting capacity for more than 1 million Maia 300 chips — though component supply constraints and ongoing capacity negotiations with TSMC could limit how quickly that goal is reached.

Courting Major Cloud Customers

Beyond ramping up production, Microsoft is also reportedly working to persuade major cloud customers to adopt the chip for their own AI workloads, including Anthropic. Microsoft has previously held early discussions about supplying its Maia 200 chips to Anthropic, though no deal had been finalized at that stage. Responding to the reporting, Andrew Wall, general manager for Microsoft's Azure Maia program, said in a statement that the company "continues to invest in custom silicon as part of our long-term AI infrastructure strategy," adding that the reported production figures "don't reflect the scale of our program."

How Maia 300 Fits Into Microsoft's Chip Roadmap

Microsoft unveiled its second-generation Maia 200 chip in January, built by TSMC using a 3-nanometer manufacturing process and packed with a significant amount of SRAM memory to help speed up handling of large volumes of user requests. The Maia 300, developed under the codename "Clea," represents the company's third attempt at proving that in-house AI silicon can work economically at scale, following limited-volume deployment of the original Maia 100 and a delayed, narrowly deployed Maia 200.

Notably, some reporting has indicated Microsoft is exploring having the Maia 300 manufactured in the United States, potentially through Intel facilities in Arizona or Ohio, rather than exclusively at TSMC in Taiwan — a move framed as a way to secure the supply chain against potential disruptions tied to geopolitical tensions around Taiwan, and to work around TSMC's typically full order books from customers like Apple and Nvidia.

The Bigger Competitive Picture

Microsoft's custom silicon push sits alongside similar efforts from Google and Amazon, both of which are racing to reduce dependence on Nvidia by developing their own AI accelerators. Google's TPU program is generally viewed as the most mature of the three, having gone through multiple generations and reportedly powering a meaningful share of Google's internal AI workloads. Analysts note that the real test for Maia 300 won't be the unveiling itself, but whether external customers with genuine flexibility to choose their hardware actually adopt it over Nvidia's chips when they aren't required to.

It's worth noting that key details — including the exact September timing, the 300,000-unit figure, and the 2027 delivery date — currently rest on reporting from The Information rather than an official Microsoft announcement. No public statement from Microsoft has yet confirmed a launch date, manufacturing node, memory specifications, or performance targets for the chip. Neither Microsoft nor TSMC has responded to requests for comment on the report. For Microsoft's official statements on its AI infrastructure strategy, see the Microsoft Azure Blog.

What to Watch Next

All eyes will be on whether Microsoft confirms a September unveiling and, if so, what concrete performance and adoption commitments accompany the announcement — particularly from major external cloud customers who could validate the chip's real-world competitiveness against Nvidia's dominant AI hardware.