TensorFlow 库未编译为使用 SSE4.1、SSE4.2、AVX、AVX2 指令来加速 CPU 计算答案

【问题标题】：The TensorFlow library wasn't compiled to use SSE4.1, SSE4.2, AVX, AVX2 instructions to speed up CPU computationsTensorFlow 库未编译为使用 SSE4.1、SSE4.2、AVX、AVX2 指令来加速 CPU 计算
【发布时间】：2020-07-23 04:45:33
【问题描述】：

我只有 Intel(R) Xeon(R) CPU E5-2640 v4 @ 2.40GHz 型号的 CPU 机器。我需要加快运行 tensorflow 代码的运行时间。在运行代码时，它会打印警告消息，例如：

    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.
    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.
    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.
    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.
    tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX512F instructions, but these are available on your machine and could speed up CPU computations.
    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.

我安装了我的 python 和 tensorflow，如下：

conda create -n  py27tf121 python=2.7
pip install tensorflow=1.2.1

我认为这个警告可以加快进程，但我不知道如何激活这些标志

【问题讨论】：

标签： linux tensorflow time cpu

【解决方案1】：

此消息意味着，如果您从源代码编译 TensorFlow，您可能会获得更好的性能。话虽如此，从源代码编译需要时间和精力。如果您想尝试这样做，您可能会发现this 很有用

【讨论】：