Energy-efficient AI processing with proven ultra-low-power microcontrollers

03-06-2021 | RS Components | Semiconductors

The Maxim MAX78000 is a new breed of AI microcontroller built to facilitate neural networks to execute at ultra-low power and live at the edge of the IoT. This product, available now from RS Components, combines the most energy-efficient AI processing with its proven ultra-low-power microcontrollers. The hardware-based convolutional neural network accelerator allows battery-powered applications to execute AI inferences while consuming only microjoules of energy.

The device is an advanced SoC featuring an Arm Cortex-M4 with FPU CPU for efficient system control with an ultra-low-power deep neural network accelerator. The CNN engine offers a weight storage memory of 442KB, and can support 1, 2, 4, and 8-bit weights (supporting networks of up to 3.5 million weights). The CNN weight memory is SRAM-based so that AI network updates may be made on the fly. The CNN engine also offers 512KB of data memory. The CNN architecture is extremely flexible, enabling networks to be trained in conventional toolsets like PyTorch and TensorFlow, then converted for execution on the device utilising tools provided.

Adding to the memory in the CNN engine, the device provides large on-chip system memory for the microcontroller core, with 512KB flash and up to 128KB SRAM. Multiple high-speed and low-power communications interfaces are supported, including I2S and a parallel camera interface.

The device is provided in 81-pin CTBGA (8mm x 8mm, 0.8mm pitch) and 130-pin WLP (4.6mm x 3.7mm, 0.35mm pitch) packages.

Typical application areas include object detection and classification; audio processing: multi-keyword recognition, sound classification, noise cancellation; facial recognition; and time-series data processing: heart rate/health signal analysis, multi-sensor analysis, predictive maintenance.

By Natasha Shek