MIPS Launches Three Workload-Native Platforms to Accelerate Physical AI

In September 2026, MIPS launched three workload-native platforms that bring Physical AI capabilities to industrial machines, transportation platforms, and embedded systems operating in the physical world. These new platforms transform commercial and ecosystem development by reducing modern AI to tailored platforms with class-leading energy efficiency.

Meet the New MIPS Platforms

As intelligence shifts to machines and vehicles, and other physical systems, compute platforms must adapt, emphasized Sameer Wasson, CEO of MIPS. MIPS is leading this change with open, software-first platforms that enable efficient, customizable Physical AI solutions by building, customizing, and deploying Physical AI. 

MIPS prioritizes software in its design approach, leveraging open-standard technology and advanced modeling to guide developers and platform designers. Their expertise in safety, real-time systems, and AI inference supports tailored solutions.

The growth of RISC-V is fueled by industry collaboration and an expanding software ecosystem,” emphasized Manuel Uhm from AMD. Further adding that they support initiatives that increase developer access and enhance software readiness for open-standard architectures.

David Ku from MediaTek emphasized that RISC-V's advancement relies on strong collaboration among processor technology, software, and the development ecosystem. MIPS is committed to open standards and enhancing RISC-V solutions with proven expertise in processor and software technology.

Impact on the Automotive Industry

These three newly launched workload-native platforms positively impact the automotive industry by accelerating the development of Advanced Driver Assistance Systems (ADAS) and autonomous driving functionalities. As the automotive industry shifts toward Software-Defined Vehicles (SDVs), tailored hardware for vehicular intelligence helps OEMs scale efficiently. 

The Inference-Native (Acies) and Safety-Native (Aegis) platforms offer complex AI algorithm processing at the vehicle level, independent of cloud systems. The Event-Native (Actus) platform ensures real-time robotic reactions, like engine management and braking, allowing vehicles to act swiftly based on sensor data.

MIPS addresses key limitations in the automotive industry, including high power consumption, data movement inefficiencies, and vendor lock-in. Its multi-threaded architecture improves data throughput by 40%, optimizes power consumption, and enables open RISC-V architectures for customizability. The platforms also improve AI stack utilization by over 30%, streamline safety validation, and mitigate hardware-software co-development timelines with tools like MIPS Atlas Explorer.

Impact on the Semiconductor and Electronic industry

This launch positively impacts the industry, where the semiconductor industry has a greater contribution in addressing the challenges of carbon taxes and decarburization. MIPS also addresses power and efficiency challenges by optimizing instructions and localized processing to mitigate data transfer. Additionally, its event-native architectures enable ultra-low latency for real-time responses, while efficiently offering modular developer platforms to mitigate customization costs for application-specific integrated circuits (ASICs).

This launch heavily benefits the semiconductor and electronics industry by accelerating the shift-left methodology enabled by systems like the MIPS Atlas Explorer by allowing chip designers to model software performance in a digital twin environment. Furthermore, MIPS establishes a standardized foundation for embedding edge AI into heavy machinery and transportation for a better future of the industry.

About MIPS

MIPS, by GlobalFoundries, develops AI inference platforms for energy-efficient Physical AI systems and embedded devices.  It also combines GF’s Physical AI portfolio with MIPS’ RISC-V and software-to-silicon expertise to accelerate custom, software-first products for automotive, industrial, and agentic edge platforms and deliver open technology. 

Furthermore, MIPS provides a single portfolio spanning AI & processor IP, advanced workload modeling, and custom solutions for Physical AI platforms. Overall, it emphasized the delivery of highly efficient, multi-threaded, real-time, and safe compute subsystems for autonomous edge platforms.