Analog Devices Inc. MAX78000 AI MCU w/ Neural Network Accelerator

Analog Devices Inc. MAX78000 AI MCU with Neural Network Accelerator is used for Artificial intelligence (AI) applications that require extreme computational horsepower. The MAX78000 enables neural networks and combines energy-efficient AI processing with ultra-low-power microcontrollers. The hardware-based Convolutional Neural Network (CNN) accelerator enables battery-powered applications to execute AI inferences while spending only microjoules of energy.

Analog Devices MAX78000 is available in an 81-pin CTBGA package with a -40°C to +105°C operating temperature range.

Features

  • Dual-core ultra-low-power microcontroller
    • Arm® Cortex-M4 processor with FPU up to 100MHz
    • 512KB Flash and 128KB SRAM
    • Optimized performance with 16KB instruction cache
    • Optional Error Correction Code (ECC-SEC-DED) for SRAM
    • 32-bit RISC-V coprocessor up to 60MHz
    • Up to 52x general-purpose I/O pins
    • 12-bit parallel camera interface
    • 1x I2S master/slave for digital audio interface
  • Neural network accelerator
    • Highly optimized for deep convolutional neural networks
    • 442k 8-bit weight capacity with 1, 2, 4, 8-bit weights
    • Programmable input image size up to 1024 x 1024 pixels
    • Programmable network depth up to 64x layers
    • Programmable per layer network channel widths up to 1024x channels
    • 1- and 2-dimensional convolution processing
    • Streaming mode
    • Flexibility to support other network types, including MLP and recurrent neural networks
  • Power management maximizes operating time for battery applications
    • Integrated Single-Inductor Multiple-Output (SIMO) Switch-Mode Power Supply (SMPS)
    • 2.0V to 3.6V SIMO supply voltage range
    • Dynamic voltage scaling minimizes active core power consumption
    • 22.2μA/MHz with loop execution at 3.0V from cache (CM4 only)
    • Selectable SRAM retention in low-power modes with Real-Time Clock (RTC) enabled
  • Security and integrity
    • Available secure boot
    • AES 128/192/256 hardware acceleration engine
    • True Random Number Generator (TRNG) seed generator
  • Package options
    • CTBGA81 (8.0mm x 8.0mm, 0.8mm pitch)
    • WLP130 (4.6mm x 3.7mm, 0.35mm pitch)

Applications

  • Object detection and classification
  • Audio processing: multi-keyword recognition, sound classification, noise cancellation
  • Facial recognition
  • Time-series data processing: heart rate/health signal analysis, multi-sensor analysis, predictive maintenance

Videos

Block Diagram

Block Diagram - Analog Devices Inc. MAX78000 AI MCU w/ Neural Network Accelerator
Publicerad: 2020-08-19 | Uppdaterad: 2025-04-23