Low Power Neural Network Accelerated Microcontroller for Battery Powered AI at Edge IoT Applications

MAX78000 Neural Network Accelerated MCU

Maxim Integrated has introduced MAX78000, a low-power neural network accelerated microcontroller that moves artificial intelligence (AI) to the edge without performance compromises in battery-powered internet of things (IoT) devices. The MAX78000 executes inferences 100x faster than software solutions running on low power microcontrollers, at a fraction of the cost of FPGA or GPU solutions.

 

Integrating a dedicated neural network accelerator with a pair of microcontroller cores, MAX78000 can enable the AI devices to see and hear complex patterns with local low-power AI processing that executes in real-time. The MAX78000 executes interferences at less than 1/100th of energy required by a microcontroller. This provides more efficient applications such as machine vision, audio, and facial recognition.

 

The core hardware of the MAX7800 minimizes the energy consumption and latency of convolutional neural networks and it runs with minimal intervention from any microcontroller core for an extremely streamlined operation. Users can also get the comprehensive MAX7800EVKIT that comes with audio and camera inputs and out-of-the-box running demos.

 

Key Features of MAX78000 Neural Network Accelerated MCU

  • Arm Cortex-M4 Processor with FPU Up to 100MHz
  • 32-Bit RISC-V Coprocessor up to 60MHz
  • Up to 52 General-Purpose I/O Pins
  • 442k 8bit Weight Capacity with 1, 2, 4, 8-bit Weights
  • Programmable Input Image Size up to 1024 x 1024 pixels
  • Programmable per Layer Network Channel Widths up to 1024 Channels
  • Supply Voltage Range: 2.0V to 3.6V 
  • 22.2μA/MHz While Loop Execution at 3.0V from Cache

 

The MAX78000 is available for purchase from authorized distributors and the MAX7800EVKIT# evaluation kit is also available for purchase at $168. To know more about the MAX78000, visit the official website of Maxim Integrated.

Component Datasheet