MIPS launches a workload-native platform to fully empower physical AI

MIPS (a subsidiary of GlobalFoundries) has announced the launch of three workload-native platforms that bring physical AI capabilities to industrial machinery, transport platforms and embedded systems operating in the physical world. These new platforms will drive commercial and ecosystem development, facilitating the deployment of modern AI on customised platforms whilst delivering industry-leading energy efficiency.

Sameer Wasson, CEO of MIPS, stated: “As intelligence extends to machines, vehicles and other physical systems, computing platforms need to adapt to the workloads. MIPS is driving this transformation with an open, software-first platform, providing a universal foundation for customers to build, customise and deploy physical AI, whilst energy efficiency and tokens per watt are the cornerstones of our technology.”

MIPS’s strategy adopts a software-first approach to design, providing guidance to developers and platform designers through open-standard technologies and advanced modelling methods. The entire process—from workload insights to the implementation of optimised platforms—benefits from MIPS’s open, modular technology and its accumulated expertise in security, real-time processing and AI inference within the field of embedded custom solutions.

Manuel Uhm, Director of Chip Marketing at AMD’s Embedded Business Unit, stated: “The continued momentum of RISC-V remains driven by strong industry collaboration and an ever-expanding software ecosystem. We welcome collective efforts to broaden developer engagement, enhance software readiness, and help customers fully leverage open-standard architectures across a wide range of application scenarios.”

Gu Dawei, Co-Chief Operating Officer and Chief Financial Officer of MediaTek, said: “The continued growth of RISC-V depends on close collaboration between processor technology, software and the broader development ecosystem. MediaTek welcomes MIPS’s commitment to open standards and the mature processor and software expertise it brings to advancing the development of commercial-grade RISC-V solutions.”