Install Tensorflow on a Jetson TX2. Helpful build scripts.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


April 1, 2017 Install TensorFlow v1.0.1 on NVIDIA Jetson TX2 Development Kit

Jetson TX2 is flashed with JetPack 3.0 which installs:

  • L4T 27.1 an Ubuntu 16.04 64-bit variant (aarch64)
  • CUDA 8.0
  • cuDNN 6


Before installing TensorFlow, a swap file should be created (minimum of 8GB recommended). The Jetson TX2 does not have enough physical memory to compile TensorFlow. The swap file may be located on the internal eMMC, and may be removed after the build.

Note: This procedure was derived from these discussion threads:

TensorFlow should be built in the following order:

Installs Java and other dependencies needed. Also builds:


Builds version 0.4.5. Includes patches for compiling under aarch64.

Git clones v1.0.1 from the TensorFlow repository and patches the source code for aarch64

Sets up the TensorFlow environment variables. This script will ask for the default python library path. There are many settings to chose from, the script picks the usual suspects. Uses python 2.7.

Builds TensorFlow.

Once TensorFlow has finished building, this script may be used to create a 'wheel' file, a package for installing with Python. The wheel file will be in the $HOME directory.

Install wheel file

$ pip install $HOME/wheel file

Build Issues

For various reasons, the build may fail. The 'debug' folder contains a version of the script which is more verbose in the way that it describes both what it is doing and errors it encounters. See the debug directory for more details.