New MCUs enable edge AI and industry-leading real-time control

Texas Instruments has released two new series of real-time MCUs, on display at electronica 2024, that deliver advancements to assist engineers in achieving more intelligent and secure processing in automotive and industrial applications. Its TMS320F28P55x series of C2000 MCUs is the industry's first portfolio of real-time microcontrollers with an integrated NPU, allowing high accuracy and low latency fault detections. The F29H85x series is built on the company's new 64-bit C29 digital signal processor core and delivers an advanced architecture with integrated safety and security features.

"Engineers increasingly strive to design automotive and industrial applications that are more energy efficient and can make faster decisions, driving a need for more scalable processing power, expanded memory, and on-chip safety and security enablers," said Amichai Ron, senior vice president, TI Embedded Processing. "For decades, our C2000 MCUs have been enabling manufacturers to measure and efficiently process real-world data. Today's introduction of enhanced real-time performance and edge AI in our C2000 portfolio will empower engineers to solve more complex problems and reach higher levels of system efficiency, safety and sustainability."

Engineers today are challenged to design systems that can make accurate, intelligent decisions in real-time to perform functions such as arc fault detection in solar and energy storage systems and motor-bearing fault detection for predictive maintenance. As the industry's first real-time MCUs with an integrated neural processing unit, its new C2000 TMS320F28P55x series meets these challenges by ensuring consistent performance with real-time processing.

The NPU in the TMS320F28P55x series offloads the execution of the neural network model from the main CPU, achieving five to ten times lower latency than software implementations to facilitate faster, more accurate decision-making. Further, the model that runs on the integrated NPU learns and adapts to different environments through training, helping systems achieve greater than 99% fault detection accuracy to enable more informed decision-making at the edge. The company's complete AI toolchain, which includes models that are optimised and tested for specific applications, helps engineers with any level of experience easily complete the AI model development process.

Designers increasingly seek a single MCU with more flash memory, larger computational capabilities and more integrated functionality. This need is especially evident as vehicles shift from discrete solutions to integrated solutions, where a single chip can handle various computational and housekeeping functions within the powertrain. The new C29 core, with its 64-bit architecture, delivers more than double the real-time signal-chain performance of the C28 core.

Featuring comprehensive diagnostics and error-checking mechanisms, the F29H85x series was engineered for compliance with International Organisation for Standardisation 26262 and International Electrotechnical Commission 61508 automotive and industrial safety standards, with levels up to ASIL D and SIL 3. These MCUs also provide cybersecurity capabilities, with a fully isolated hardware security module that protects systems from unauthorised access and cyberthreats. Additionally, the company's proprietary safety and security unit provides runtime safety and security without a performance penalty, employing an advanced, context-aware memory protection unit for hardware isolation of CPU tasks, with freedom from interference and self-test processes.

At electronica 2024, the is showcasing how its analog and embedded processing products are helping design a more intelligent and sustainable future.