March 17, 2026 /SemiMedia/ — Texas Instruments has introduced two new microcontroller families with built-in neural processing units (NPUs), aiming to bring edge artificial intelligence (AI) capabilities into a wider range of embedded systems.

The MSPM0G5187 and AM13Ex devices integrate TI’s TinyEngine NPU, allowing AI workloads to run alongside the main CPU. This setup improves processing speed and reduces power use, which is critical for small and battery-powered devices.

From a hardware perspective, the added NPU enables parallel execution of neural network tasks, reducing reliance on the CPU. This makes it possible to run AI functions directly on microcontrollers used in wearables, home electronics and industrial equipment.

The MSPM0G5187, based on an Arm Cortex-M0+ core, targets cost-sensitive applications. With hardware acceleration, it cuts latency and energy consumption during AI inference while lowering memory usage. Priced below $1 in volume, the chip is positioned as an entry-level option for adding AI features to simple devices.

For industrial use, the AM13Ex series combines an Arm Cortex-M33 core, real-time control functions and the NPU in a single chip. This allows designers to manage motor control and AI tasks at the same time without extra components, helping reduce system cost and complexity.

TI also supports both MCU families with its development tools, including an AI-focused software environment that helps engineers train and deploy models more easily. The platform includes dozens of pre-built models to speed up development.

The MSPM0G5187 is already available in production volumes, while the AM13Ex series is currently in preproduction, with more versions expected by the end of 2026.