Skip to content
BSP and examples to integrate the SparkFun TensorFlow board with the AmbiqSuite
C Python Makefile
Branch: master
Clone or download
oclyke Merge pull request #14 from suphoff/loader_clear_serial Discard buffered data before hello
Latest commit 60d4be1 Aug 26, 2019
Type Name Latest commit message Commit time
Failed to load latest commit information.
bsp Discard buffered data before hello Aug 26, 2019
examples example2_dual_mic_dma clarifications Jun 1, 2019
.gitignore Un-ignore .a files to include built libam_bsp.a Feb 22, 2019
Makefile Huge restructuring Feb 8, 2019 Update Mar 29, 2019


BSP and examples to integrate the SparkFun TensorFlow board with the AmbiqSuite SDK.

Associated Documentation


Clone or download then extract or symlink this repo into the SDK/boards/ directory.


This repo contains several example projects.

  • example1_edge_test demonstrates the hardware features of the board. The makefile is preconfigured with relative paths.
    • To compile with gcc and flash use the makefile in the gcc subdirectory.
      • Edit 'COM_PORT' variable to specify where to try UART flashing
      • make
      • make clean
      • make bootload
  • example2_dual_mic_dma shows how to use DMA to capture 16 kHz audio recordings from both microphones without blocking the processor. This is the basis of how audio is provided to the TensorFlow Lite demo.
  • tensorflow_demo uses a pre-trained model to identify "yes" and "no" and blink a corresponding LED on the board. Using GPIO you could easily expand this example to control a device.
  • SparkFun_Edge_Project_Template has a relatively easy to set up makefile with some example header files and source files included. You can copy this directory to an arbitraty location on your filesystem to begin a new project.
    • You must provide the absolute path to the SDK root directory in the 'SDKPATH' variable
    • Also update the COM_PORT variable to match your setup
You can’t perform that action at this time.