MIPS, a subsidiary of GlobalFoundries, has partnered with Inova Semiconductors GmbH to develop a robotics control reference platform aimed at accelerating the development of advanced humanoid robots and physical artificial intelligence (AI) edge systems.
The companies said the platform is designed to address growing demand for scalable compute architectures capable of supporting real-time robotic control alongside secure AI processing. By combining high-speed communication, open compute architectures, and low-power semiconductor technologies, the initiative seeks to simplify the design and deployment of next-generation robotic systems.
The reference platform integrates Inova’s experience in automotive zonal architectures—an approach commonly used in modern vehicle electronics to centralize control functions—with compute technologies based on the RISC-V instruction set architecture. The architecture is intended to support mixed-criticality workloads, enabling real-time control loops while simultaneously handling AI-driven decision-making processes.
At the hardware level, the platform combines several technologies from MIPS’ Atlas processor portfolio, including the MIPS Atlas M8500 high-performance microcontroller processor IP and the MIPS Atlas S8200 AI processor IP. These components are paired with Inova’s APXpress high-speed interface technology to enable efficient data aggregation and communication across multiple robotic subsystems.
Manufacturing for the platform will utilize the FDX semiconductor process technology from GlobalFoundries. The process is designed to provide energy-efficient performance, which is particularly important for robotics systems operating at the edge where power constraints are a key consideration.
The companies said the platform will support real-time, high-performance multi-axis motion control, a requirement for advanced robotics applications such as humanoid systems and robotic arms. By integrating sensing, processing, and control capabilities into a unified architecture, the platform is expected to reduce development complexity while enabling faster prototyping and product scaling.
Sameer Wasson, CEO of MIPS, said the collaboration aims to provide robotics developers with an open and scalable hardware foundation. He noted that integrating high-speed connectivity with RISC-V-based computing could help reduce system costs while accelerating development timelines.
Robert Isele, CEO of Inova Semiconductors, said the approach builds on architectural concepts widely used in automotive electronics to address similar challenges in robotics. According to Isele, a zonal architecture could provide robot developers with a standardized framework that simplifies system integration while enabling secure and deterministic connectivity.
Early development access to the platform will be provided through the MIPS Atlas Explorer simulation environment, which allows software developers to test and optimize applications using virtual models of the platform’s compute components before hardware deployment.
The companies said the collaboration reflects the increasing convergence of robotics, AI, and semiconductor technologies as developers seek platforms capable of supporting more sophisticated physical AI systems.
