Microsoft unveiled on October 12, 2026 the Azure Maia 300, its third-generation custom AI inference chip. Built on TSMC’s N2P process node using chiplet architecture, the chip delivers 40% more tokens per megawatt than current leading GPU solutions and up to 7x the throughput of its predecessor, the Maia 200. Deployment at scale in Azure data centers is planned to begin in 2027. Microsoft’s capital expenditure exceeds $175 billion annually, yet the Maia 300 is expected to account for only around 10% of the company’s AI chip spending by early 2027. This custom silicon initiative aims to generate multi-billion-dollar savings and reduce dependence on Nvidia and AMD, particularly for inference workloads.
Source: Read the original article
Disclaimer: this content is for information purposes only and is not financial advice. Cryptocurrencies are highly volatile: you may lose all of your capital. Always do your own research. Legal notice
