Open-Source SparkFun Module Supports Low-Power TensorFlow Machine Learning
July 17, 2019
In addition to a secure firmware update system, and camera compatibility, the BLE-enabled Artemis board features large SMD pads that support three companion carrier boards.
SparkFun has released the SparkFun Artemis, Engineering Version, an open-source embedded development kit that supports the TensorFlow machine learning environment. Designed for toolchain-agnostic, low-power machine learning development, the 15.5 mm x 10.5 mm Artemis board includes:
- A Bluetooth Low Energy (BLE) module featuring the Arm Cortex-M4F-based Apollo3 MCU from Ambiq Micro, integrated Bluetooth 5 radio, and 2.4 GHz antenna
- Ability to run machine learning algorithms at 6 µA/MHz at 3.3B
- Hardware Abstraction Layer (HAL)
- Programming via serial bootloader or JTAG
- ISO7816 Secure ‘Smart Card’ interface
In addition to a secure firmware update system, flexible, serial peripherals, a suite of clock sources, and camera compatibility, the Artemis board features large SMD pads that support carrier board implementations. SparkFun has launched three carrier boards in conjunction with the release of the Artemis, Engineering version board: the BlackBoard Artemis (Arduino Uno footprint); BlackBoard Artemis Nano (smallest form factor); and BlackBoard Artemis ATP (with 48 GPIO pins).
The carrier boards each contain a BLE radio, digital MEMS microphone, JTAG and Qwicc connectors, and are compatible with the SparkFun Arduino core that can be programmed using the Arduino IDE.
he Artemis ecosystem is available now. For more information visit www.sparkfun.com/artemis.