MAX78000EXG Artificial intelligence (AI) requires extreme computing horsepower, but Maxim is cutting the cord for AI insights. The MAX78000 is a new AI microcontroller designed to enable ultra-low-power execution of neural networks, living at the edge of the Internet of Things. This product combines the most power-efficient AI processing with Maxim's proven ultra-low-power microcontrollers. Our hardware-based convolutional neural network (CNN) accelerator enables battery-powered applications to perform AI inference while consuming only microjoules of energy.

The MAX78000EXG is an advanced system-on-chip featuring an Arm® Cortex®-M4 with FPU CPU for high-efficiency system control using an ultra-low-power deep neural network accelerator. The weight storage memory of the CNN engine is 442KB, which can support 1, 2, 4, and 8-bit weights (supporting networks with up to 3.5 million weights). CNN heavy memory is SRAM based so AI network updates can be made on the fly. The CNN engine also has 512KB of data memory. The CNN architecture is highly flexible, allowing networks to be trained in traditional toolsets like PyTorch® and TensorFlow®, and then converted to execute on the MAX78000 using tools provided by Maxim.

Product parameters:

Product Category: ARM Microcontrollers - MCUs

RoHS: Details

Series: MAX78000

Installation style: SMD/SMT

Core: ARM Cortex M4

Program memory size: 512 kB

Data bus width: 32 bit

ADC resolution: 10bit

Maximum clock frequency: 100 MHz

Number of inputs/outputs: 52 I/O

Data RAM size: 512 kB

Supply voltage - Min: 2 V

Supply Voltage - Max: 3.6 V

Minimum working temperature: - 40 C

Maximum working temperature: + 105 C

Data Ram Type: SRAM

Interface type: I2C, I2S, GPIO, PWM, SPI, UART

Humidity Sensitivity: Yes

Product: MCU+FPU

Product Type: ARM Microcontrollers - MCU

Program memory type: Flash

Subcategory: Microcontrollers - MCU

Watchdog Timer: Watchdog Timer