CanMV-K230 is a credit card-sized development board for AI and computer vision applications based on the Kendryte K230 dual-core C908 64-bit RISC-V processor with built-in KPU (Knowledge Process Unit) and various interfaces such as MIPI CSI inputs and Ethernet.
The first Kendryte RISC-V AI processor was launched in 2018 with the K210 which I tested with the Grove AI HAT and Maixduino board and found fun to experiment with, but noted that performance was limited. Since then the company introduced the K510 mid-range AI processor with a more powerful 3 TOPS AI accelerator, and the K230 entry-level successor to the K210 – which was planned for 2022 in a 2021 roadmap – has now just been launched and integrated into the CanMV-K230 development board.
CanMV-K230 specifications:
- SoC – Kendryte K230
- CPU
- 64-bit RISC-V processor @ 1.6GHz with RISC-V Vector Extension 1.0, FPU
- 64-bit RISC-V processor @ 800MHz with support for RV64GCB instruction set;
- AI accelerator
- KPU with support for INT8 and INT16
- Compatible with TVM, TensorFlow, Pytorch, ONNX
- Typical network performance: Resnet 50 ≥ 85fps @ INT8; Mobilenet_v2 ≥ 670fps @ INT8; YoloV5S ≥ 38fps @INT8;
- “DPU” using 3D structured light depth calculation up to 1280×800 @ 30fps
- VPU – Video encoding/decoding for H.264/H.265/JPEG/MJPEG up to 4Kp40/4Kp20
- CPU
- System Memory: 512MB LPDDR3
- Storage – QSPI flash, microSD card slot
- Display – HDMI port, MIPI DSI connector up to 1080p60
- Audio – 3.5mm audio jack
- Camera – Up to 5MP camera, up to 3-ch MIPI CSI inputs
- Networking
- 10/100Mbps Ethernet RJ45 port
- WiFi 4 and Bluetooth 4.0 via AP6212 wireless module
- USB – 1x USB 2.0 OTG Type-C port port
- Expansion – 40-pin GPIO header with up to 29x GPIOs, 5x PWM, 4x I2C, 2x UART
- Misc – Buttons, RGB LED
- Power Supply – 5V via USB-C port
- Dimension – 85 x 56 mm
You’ll find the SDK and documentation on GitHub, but while there are folders for English (en) and Chinese (zh), the former only includes a product brief with some block diagram, and the vast majority of the software and hardware documentation is only available in Chinese at this time. If that is an issue, you may want to wait until a company that caters to the overseas market (e.g. outside of China) makes a K230 board. Having said that, I can see over 30 AI demos from face detection to license plate recognition or smoke detection to play with and it might not be too hard to get started even if you can’t read Chinese, but for more advanced use cases I’d probably wait until the English documentation is ready. The system appears to run two operating systems: RT Smart RTOS and another based on Linux, while the SDK is compatible with Ubuntu 20.04 or available in a Docker image.
While the company lists the K510 as a mid-range chip and the K230 as an entry-level device in the 2021 roadmap, the latter delivers higher efficiency with Mobilenet V1, Resnet 50, and YoloV5S with up to 341 FPS/TOPS against 133 FPS/TOPS for the K510 and 86 FPS/TOPS the original K210 SoC thanks to improved MAC utilization.
The CanMV-K230 development board with the Kendryte K230 AI RISC-V SoC can be purchased on AnalogLamb for $49.99.
Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.
Support CNX Software! Donate via cryptocurrencies, become a Patron on Patreon, or purchase goods on Amazon or Aliexpress
Already got this interesting board! A RTOS is running on the big RVV1.0 equipped core. It comes with a IoT tailored BSP, so porting Linux to run on the “big core” might need so extra effort.
how many TOPS is the k230?
Do anyone knows an estimated price of the chip?
thank you!