Intel’s 18A process node, developed in-house at its Fab 52 semiconductor facility in Arizona, is positioned for both internal and external deployments. Intel says the production-ready technology will power its upcoming Panther Lake and Clearwater Forest chips, while Intel Foundry is open for business to partners targeting 18A or the enhanced 18A-P variant. The strategy reflects Intel’s push to re-establish itself as a leading manufacturing option for high-performance semiconductors.
According to reporting from SemiAccurate, cited by TechPowerUp, Microsoft is planning to use Intel’s 18A process for its next-generation Maia 3 accelerator for Artificial Intelligence, codenamed “Griffin.” The report adds that, depending on yields and the success of this latest hardware, Microsoft could also consider future Intel nodes like 18A-PT and 14A for subsequent Maia accelerators. Concrete specifications for Maia 3 remain scarce, but the sourcing suggests Microsoft is preparing a significant manufacturing partnership around its accelerator roadmap.
Microsoft’s interest in Intel Foundry is not new. When Intel outlined its Foundry roadmap in early 2024, Microsoft CEO Satya Nadella said the company needed a reliable supply of advanced, high-performance, and high-quality semiconductors, and that Microsoft had chosen a chip design it planned to produce on Intel’s 18A process. That statement signaled intent to diversify manufacturing for critical cloud and Artificial Intelligence infrastructure.
The latest report also notes Microsoft could opt for the more advanced 18A-P process, which combines RibbonFET and PowerVia technologies to improve performance and efficiency. For Intel, winning high-profile customers is central to making 18A and Intel Foundry a viable alternative in the market. Other notable customers and partners cited for Intel’s next-gen manufacturing include NVIDIA, Broadcom, and ARM, underscoring growing industry interest as advanced nodes come online.
