TI brings edge AI microcontrollers into embedded design

Texas Instruments has recently announced the expansion of its embedded portfolio with the introduction of two new MCU families, aimed at advancing edge AI microcontrollers into lower-cost and more control-intensive designs. These new families combine dedicated neural acceleration with a broader software push focused on model deployment and AI-assisted development. The launch is centered around the MSPM0G5187, a general-purpose Arm Cortex-M0+ device, and the AM13Ex family, which is designed for higher-performance real-time control.

Both MCU families integrate TI’s TinyEngine neural processing unit, which is intended to offload inference work from the main CPU, making local AI practical in systems that previously relied solely on standard MCU compute power. According to TI, the TinyEngine block can significantly reduce latency by up to 92 times and energy consumption per inference by up to 123 times compared to similar MCUs without a dedicated accelerator.

For the MSPM0G5187, this advancement is crucial as it enables AI to be integrated into wearables, appliances, sensor nodes, and electrical control products, expanding the reach of AI capabilities beyond high-end SoCs and processor-based designs. Moreover, TI emphasizes affordability by pricing the MSPM0G5187 at under US$1 in 1,000-unit quantities, signaling that edge AI microcontrollers are now accessible for a wider range of designs.

On the higher-performance side, the AM13Ex family combines an Arm Cortex-M33 core, the TinyEngine NPU, and real-time control hardware on a single chip. This configuration is targeted at applications in appliances, robotics, and industrial systems where adaptive control, condition monitoring, and predictive maintenance need to coexist with deterministic control loops.

TI highlights that the AM13Ex can effectively manage real-time control for up to four motors while simultaneously running AI-based adaptive control algorithms. Additionally, the integrated trigonometric maths accelerator in the AM13Ex performs calculations 10 times faster than traditional CORDIC implementations, contributing to a potential bill-of-materials cost reduction of up to 30% by eliminating the need for extra external components.

Aside from the hardware advancements, TI is also enhancing its software tools ecosystem. The company is pairing the new MCU families with CCStudio Edge AI Studio, which now offers over 60 models and application examples. Furthermore, TI has integrated generative AI functions into the CCStudio IDE for code development, configuration, and debugging, providing developers with a comprehensive set of tools to streamline the development process.