SparkFun Artemis Module - Low Power Machine Learning BLE Cortex-M4F

The SparkFun Artemis Module is a compact 10 × 15 mm wireless module built around the Ambiq Apollo3 Cortex-M4F processor with BLE 5.0. Running at 48 MHz (with 96 MHz turbo mode), it draws as little as 6 µA per MHz — making it exceptionally power-efficient for battery-powered and always-on applications including machine learning inference at the edge.

The module includes all necessary support circuitry for the Apollo3 processor, and its open-source hardware design can be implemented on low-cost 2-layer PCBs with 8 mil trace/space. Program it using the Arduino core or push further with the advanced HAL for full access to the Cortex-M4F architecture.

Key Features

• Cortex-M4F Processor – 48 MHz default, 96 MHz turbo mode
• Ultra-Low Power – 6 µA/MHz (<5 mW), ideal for battery-powered projects
• BLE 5.0 – Built-in Bluetooth Low Energy
• 1 MB Flash + 384 KB RAM – Ample space for complex applications
• Arduino Compatible – Program using the Arduino IDE and SparkFun's Arduino core
• Compact Module – 10 × 15 mm with all support circuitry included
• Open Source Hardware – Full design files available for custom carrier board development
• Certified – FCC, IC, and CE certified; available in tape and reel

Specifications

• Processor: Ambiq Apollo3 (ARM Cortex-M4F)
• Clock Speed: 48 MHz (96 MHz turbo)
• Flash: 1 MB
• RAM: 384 KB
• Wireless: BLE 5.0
• Power: 6 µA/MHz
• Module Size: 10 × 15 mm
• Certifications: FCC, IC, CE

Ideal For

• Ultra-low-power IoT devices
• Edge machine learning and voice recognition
• BLE-connected wearables and sensors
• Custom product development and production scaling

Resources

• Arduino Core (GitHub)
• Hardware Design Files (GitHub)