Skip to content
A TensorFlow 2.0 .whl file compiled with an old processor/computer
Branch: master
Clone or download
Latest commit 6cf5cdd Dec 7, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
README.md

README.md

Tensorflow2.0ForOldComputers for Linux

A TensorFlow 2.0 .whl file built with an old processor/computer (for Linux).

Old Computer

As this file is too big for GitHub I upload it on MediaFire.

Follow this link : http://www.mediafire.com/file/w4iduu2ypqefd98/tensorflow-2.0.0-cp36-cp36m-linux_x86_64.whl/file#

As you may know, since TensorFlow 1.6, official binaries use AVX instructions which may not run on older CPUs. I compiled this 2.0 version under Ubuntu 18.04.3 LTS with a HP Pavillon dv6 notebook running with an Intel Core 2 duo T6600 @2.20GHZ x 2 which is an old CPU (2009) not compatible with AVX instructions nor SSE.

See Technical Specifications for this CPU on Intel WebSite : https://ark.intel.com/content/www/us/en/ark/products/37255/intel-core-2-duo-processor-t6600-2m-cache-2-20-ghz-800-mhz-fsb.html

(you can check your own CPU on this site)

Logo Tensorflow

Download the file in a directory for example /tmp/tsf :

/>cd tmp

/tmp>mkdir tsf

/tmp>cd tsf

/tmp/tsf>wget http://www.mediafire.com/file/w4iduu2ypqefd98/tensorflow-2.0.0-cp36-cp36m-linux_x86_64.whl

Then install the .whl file as a standard library in your Python env :

/tmp/tsf>pip install tensorflow-2.0.0-cp36-cp36m-linux_x86_64.whl

If you want to install this tensorflow in a specific python/Anaconda environment, please don't forget to activate it before :

(base) /tmp/tsf>conda activate myenv

(myenv) /tmp/tsf>pip install tensorflow-2.0.0-cp36-cp36m-linux_x86_64.whl

You can’t perform that action at this time.