Microchip Technology has expanded its dsPIC33A digital signal controller (DSC) portfolio with two new families: the dsPIC33AK512MPS512 and dsPIC33AK512MC510. These latest entries target fast-growing AI Edge applications across data centres, industrial control, and automotive ECUs. The launch brings precision control, integrated security, and onboard ML inferencing to the edge. This helps in making them ideal for real-time embedded intelligence.
The new DSCs are tailored for use in power conversion, motor drives, and secure embedded platforms that require real-time responsiveness and robust machine learning integration. With this move, Microchip is doubling down on compact, secure, AI-ready controllers designed for critical environments.
Precision Control Built for AI Edge Workloads
The dsPIC33AK512MPS512 family delivers ultra-precise 78-picosecond PWM resolution and 40Msps ADC sampling speeds. Features that enable high-speed feedback loops are necessary for AI-driven control systems. This makes them ideal for use in AI Edge workloads where real-time response to data streams is critical. The series supports high-end power stage designs like SiC and GaN DC–DC converters, allowing AI-enabled systems to manage power more efficiently and securely.
For cost-conscious applications, the dsPIC33AK512MC510 series offers 1.25-nanosecond PWM resolution while retaining the same fast ADC capabilities. These specifications allow the DSCs to maintain the responsiveness required for real-time ML inferencing in motor control or embedded monitoring systems.
ML Inferencing Support Native to Both DSC Families
Microchip’s new DSCs are not just fast—they’re intelligent. Both the MPS512 and MC510 families offer native support for ML inferencing at the edge. Using the MPLAB Machine Learning Development Suite, developers can integrate data prep, model training, validation, and deployment directly into the firmware layer. These tools remove the friction typically associated with bringing machine learning to microcontroller-level systems.
Each chip features a 200MHz core, double-precision floating-point unit, and single-cycle MAC operations—optimised to handle ML inferencing loops with minimal latency. These specs allow AI models to operate locally on the edge device. This helps in reducing dependency on cloud compute and enhancing speed, privacy, and security.
Security-First Design with Trusted Features
Security remains a central theme in Microchip’s AI strategy. Both DSC lines feature Secure Boot, trusted flash, and crypto accelerators. These features offer hardware-backed protection for mission-critical artificial intelligence (AI) applications in industrial and automotive spaces.
The new controllers are also ISO 26262 and IEC 61508 compliant, making them eligible for AI applications within safety-certified systems. However, secure Boot prevents tampering when the device starts up, while the integrated cryptographic modules ensure sensitive/valuable ML models are not exposed or manipulated. Microchip combined ML inferencing with Secure Boot so intelligent edge devices, in addition to being smart, are now resilient to attacks from the cyber domain and functional domain.
Developer Ecosystem Built for AI Acceleration
Microchip’s new DSCs are fully supported in the MPLAB ecosystem. This allows developers the ability to quickly prototype and deploy AI models to the cloud. Provisioning their DSC on SAFERTOS and incorporating the TRACE32 debug interface. The developers can accelerate their development of Embedded AI systems in safety-critical areas. The integrated support for ML tools and highly configurable debug options. This allows developer teams to focus more on model performance and less on the systems integration. Starting at price points of $1.50 (depending on the mlnired configuration), these AI-capable DSCs provide serious capability at a competitive price.
About The Author
