English
Two years ago, STMicroelectronics launched the STMP32MP25 , its first offering in the STMP32MP2 microprocessor (MPU) family . Now, the company is back with a new offering, the STMP32MP23 MPUs series for ruggedized industrial and edge AI systems.
The STMP32MP23 MPUs are the newest addition to the STMP32MP2 family.
Modern edge deployments must balance several tradeoffs, including security, compliance, performance, and reliability. ST designed the STM32MP2 to check each of these boxes with a slew of hardware features and development tools that make design easy and straightforward.
The Newest Variants to the STMP32MP2 MPUs
ST's STM32MP2 series addresses real-time tasks in mixed-criticality systems. To this end, each member of the family integrates dual 64-bit Arm Cortex-A35 cores at 1.5 GHz and a 400-MHz Arm Cortex-M33 core to respond to various tasks.
Each family member offers distinct features. For example, the STM32MP257, MP255, and MP235 variants incorporate a 0.6 TOPS AI neural accelerator, a 3D GPU, and H.264 decode capability. These devices also include support for LVDS and DSI display interfaces to support graphical HMI designs and machine vision systems.
The STM32MP2 family portfolio.
The STM32MP253 and MP251 versions focus on industrial control and edge processing, with the MP253 offering three FD-CAN interfaces for distributed sensor and actuator networks. The MP251 features a single Ethernet port and reduced peripheral count to support cost-sensitive applications.
The new MP233 and MP231 variants offer baseline configurations for general-purpose processing, with dual and single Ethernet port options, respectively. Each device supports industrial-grade operation from -40°C to 125°C and optional hardware cryptographic features. The product line also targets PSA Level 3 and SESIP3 certification readiness with TrustZone support, secure boot, tamper detection, and cryptographic acceleration.
A portion of the STM32MP23xC/F's block diagram.
As a whole, the devices are available in four BGA packages, including TFBGA436 (18 mm x 18 mm), TFBGA361 (16 mm x 16 mm), VFBGA424 (14 mm x 14 mm), and VFBGA361 (10 mm x 10 mm). The MPUs also share pin compatibility across the STM32MP23 and MP25 series, so system architects can scale between features and costs without redesigning the board.
Development and Software Support
The STM32MP2 series benefits from a mature and extensible software ecosystem. For example, ST announced that it is extending maintenance support for its OpenSTLinux distribution from two to five years for each of these new MPUs. As such, developers will be guaranteed ongoing access to security patches and kernel updates.
For AI development, ST provides a full pipeline, including the Edge AI Core desktop tool and the Edge AI Developer Cloud , to convert, optimize, and benchmark neural networks. The integrated AI accelerator supports frameworks such as TensorFlow Lite and ONNX, with the software stack built for deployment on the 0.6 TOPS NPU.
The STM32MP257 Discovery Kit.
Finally, developers can begin evaluation with the related STM32MP257 Discovery Kit, which ships with pre-integrated software and references for STM32MP23x-specific configurations.
Preparing for Long-Term Edge Deployments
The STM32MP2 series focuses on modularity, regulatory alignment, and upgrade-ready infrastructure. Supported by a plethora of software and development tools, the MPUs should give designers a clear-cut way to develop edge deployments that are reliable and secure for the new hardware features.
The MPUs are available now through global distribution channels.
