M1 AI+lOT Module K210 Deep learning

Name: M1 AI+lOT Module K210 Deep learning
Brand: DFRobot
SKU: DF-DFR0636
Price: 30.21 AUD
Availability: InStock

via DFRobot

$30.21

SKU DF-DFR0636

Quantity

AI chip normally refers to ASIC chip that aims at AI algorithms. Although the conventional CPU and GPU can be used to execute AI algorithms, they have been greatly limited in their speed, performance, and practicality. Compared with traditional processor chip, AI chip offers faster speed, more computing power, and low energy consumption.

This product employs AI chip K210 as the core unit. K210 comes with dual-core processors with independent FPU, 64 bits CPU bit width, 8MB on-chip SRAM, 400M adjustable nominal frequency, and double-precision FPU supporting multiplication, division, and square root operation.

The AI Module-M1 is equipped with neural network hardware accelerator KPU, voice processing unit (APU), programmable IO array (FPIOA/IOMUX), and Fast Fourier Transform Accelerator. In the AI processing, K210 can perform operations such as convolution, batch normalization, activation, and pooling. At the same time, the pre-processing of voice direction scanning and voice data output can also be performed.

Features

CPU: RISC-V dual-core 64bit

400MHz Frequency (overclockable)

Debugging Support: high-speed UART and JTAG interface for debugging
Neural Network Processor: each layer of convolutional neural network parameter can be configured separately, including the number of input and output channels, and the input and output line width and column height.

support for 1×1 and 3×3 convolution kernels

Image Recognition: QVGA@60FPS/VGA@30FPS
Audio Processor: support up to 8 channels of audio input data, ie 4 stereo channels

16 bit wide internal audio signal processing

support for 12-bit, 16-bit, 24-bit, and 32-bit input data widths

Up to 192KHz sample rate

Static Random-Access Memory (SRAM): the SRAM is split into two parts, 6MiB of on-chip general-purpose SRAM memory and 2 MiB of on-chip AI SRAM memory
Field-Programmable IO Array: FPIOA allows users to map 255 internal functions to 48 free IOs on the chip
Digital Video Port: maximum frame size 640x480
FFT Accelerator: the FFT accelerator is a hardware implementation of the Fast Fourier Transform(FFT)
Deep Learning Frame: TensorFlow/Keras/Darknet
Peripherals: FPIOA, UART, GPIO, SPI, I2C, I2S, WDT, TIMER, RTC, etc

Specification