Intel® Distribution for Python*
Support and discussions for achieving faster Python* applications and core computational packages.
381 Discussions

IDP distro of Tensorflow reports no SSE, AVX, FMA at runtime

Joachim_P_Intel
Employee
350 Views

Hi,
This is on Ubuntu 16.04 LTS on a BDW-E machine. I am using Anaconda to switch to the IDP distributed packages.

When calling TF, both the Ubuntu distro (expected), but also the IDP channel report the following runtime warnings:

2017-12-21 09:03:06.483622: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
2017-12-21 09:03:06.483646: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.
2017-12-21 09:03:06.483651: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.
2017-12-21 09:03:06.483654: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.
2017-12-21 09:03:06.483658: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.

I assumed that the IDP distro of TF would be distributed with Intel specific optimizations, but that seems not the case?

JP

 

0 Kudos
1 Reply
DavidLiu
Employee
350 Views

Hi Joachim, 

Unfortunately we still see the runtime warnings as of last release, even though they were built correctly.  As long as you got it from our instructions, it should have the optimizations enabled.

Thanks,

David

Reply