STMicroelectronics (STM) investor relations material

Unified platform and physical AI

• Strategy centers on a unified STM32 MCU platform for connected, secure, software-defined systems, supporting cross-family software reuse, faster development, lower risk and lower total cost of ownership.

• Platform foundations: shared Arm Cortex-M architecture, STM32Cube and Arm ecosystem tools, and shared process technologies including 40 nm flash and 18 nm FDSOI with embedded phase-change memory.

• Technology building blocks combine Arm CPUs and ST neural accelerators for edge AI, deterministic Ethernet connectivity, SHA acceleration and PQC libraries for security, and qualified hardware/software for safety.

• The STM32 ecosystem serves close to 2 million developers; ST Community attracts over 300,000 monthly visitors and is adding the AI-powered STM32 Sidekick assistant.

• Humanoid robotics was described as nearing a $100bn annual opportunity; Arm cited a potential $200bn physical-AI compute market by the middle of the next decade.

Arm collaboration and physical AI ecosystem

• Arm named ST a lead partner in defining the future Cortex-M portfolio; Cadence will also serve as a key design partner. The collaboration targets software compatibility and high-performance MCU needs for physical AI.

• The Arm Total Design Program for physical AI includes over 100 companies, working on shared robotics capabilities and frameworks to accelerate adoption.

• STM32Cube will add NVIDIA Holoscan Sensor Bridge as a native software component in coming months, extending support across the vast majority of STM32 products; libraries for sensing, MEMS, MCUs and actuation are also planned for simulation environments.

• NVIDIA highlighted ST’s sensing, power management, processing, safety and security portfolio for distributed robot control; simulation can accelerate robot testing and development by orders of magnitude.

STM32A automotive and high-safety portfolio

• STM32A extends STM32 foundations into automotive and high-safety applications, with compatible IP, software concepts, tools and ecosystem continuity across vehicle and industrial designs.

• STM32A4 targets entry-level automotive nodes and compact industrial/infrastructure edge devices, including sensing, actuation, lighting, access control and safety monitoring.

• STM32A5 targets body and comfort control, with up to 4 MB memory, enhanced security and support for applications up to ASIL B; use cases include industrial automation, robotics and energy systems.

• STM32A9 targets vehicle zonal endpoints with deterministic real-time control, Ethernet-oriented connectivity and safety support up to ASIL D; split-lock cores can prioritize ASIL D or provide twice the ASIL B performance when ASIL D is not needed.

• Volkswagen Group described STM32A as an important step for future automotive electronics, citing platform choice and a unified software ecosystem.