Limited record: this part has neither specifications nor a datasheet in the catalog yet. It is shown for reference and is not indexed.

MF0 – Machine learning vision core module with dual-core RISC-V

Machine learning vision core module with dual-core RISC-V

System on Modules (SoM) & Computer on Modules (CoM)

  • RISC-V
Manufacturer part number
MF0
Manufacturer
Sipeed

Sipeed MF0 machine learning vision module with a 64-bit dual-core RISC-V processor, designed for embedded image recognition with TensorFlow, Keras and Darknet frameworks.

Product overview

Sipeed MF0. The Sipeed MF0 is a machine learning core module of the System on Module type, built around a 64-bit dual-core RISC-V processor running at 400 MHz by default and overclockable.

Performance. The module performs image recognition at QVGA@60FPS or VGA@30FPS and supports the TensorFlow, Keras and Darknet deep learning frameworks.

Memory. The module provides 4 MByte of ROM, 6 MByte of general purpose RAM and 2 MByte of RAM dedicated to the KPU.

Interfaces and peripherals. The module integrates FPIOA, UART, GPIO, SPI, I²C, I²S, WDT and TIMER peripherals, together with an onboard camera connector. It is offered in a 2.54 mm pitch DIP or SMT pad version and can be connected to a breadboard.

Power. The Vin input accepts 3.4 to 5.9 V, with 5.0 V recommended. An EN pin enables the module power supply, with a default high level enable.

Serial protocols. The module ships by default with the HEX format serial protocol firmware and also offers a JSON format serial protocol firmware, selectable according to the application. Firmware updates are performed with kflash_gui, the module being put into ISP download mode.

Integration. The module connects to a host system through its UART, SPI, I²C or I²S pins and fits Sipeed development boards and docks. It is available on quotation from Atlas Electronics.

See also

Technical questions

Ask a question

No published question about this part yet. Questions are reviewed before publication.

Request a quote