因此，我在笔记本电脑上安装了适用于 Python 3.5 的 Tensorflow，这是一台托管 Nvidia Geforce Pascal GPU 的 Windows 机器。我还安装了 CUDA 并下载了 cuDNN 并将其添加到 PATH 变量中。我的 tensorflow 代码确实可以编译，但是如果我监控我的 GPU，我可以看到，它不会计算任何东西，而是我的 CPU 正在完成整个工作。我还在控制台中得到一个输出，确认检测到 GPU：
2017-06-02 15:22:22.140283: W c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE instructions, but these are available on your machine and could speed up CPU computations.
2017-06-02 15:22:22.140600: W c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE2 instructions, but these are available on your machine and could speed up CPU computations.
2017-06-02 15:22:22.140899: W c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE3 instructions, but these are available on your machine and could speed up CPU computations.
2017-06-02 15:22:22.141108: W c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\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.
2017-06-02 15:22:22.141321: W c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\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.
2017-06-02 15:22:22.141582: W c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\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.
2017-06-02 15:22:22.141803: W c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\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.
2017-06-02 15:22:22.142130: W c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\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.
2017-06-02 15:22:22.561687: I c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\tensorflow\core\common_runtime\gpu\gpu_device.cc:887] Found device 0 with properties:
name: GeForce GTX 1070
major: 6 minor: 1 memoryClockRate (GHz) 1.645
pciBusID 0000:01:00.0
Total memory: 8.00GiB
Free memory: 6.65GiB
2017-06-02 15:22:22.561949: I c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\tensorflow\core\common_runtime\gpu\gpu_device.cc:908] DMA: 0
2017-06-02 15:22:22.562073: I c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\tensorflow\core\common_runtime\gpu\gpu_device.cc:918] 0: Y
2017-06-02 15:22:22.562435: I c:\tf_jenkins\home\workspace\release-win\device\gpu\os\windows\tensorflow\core\common_runtime\gpu\gpu_device.cc:977] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX 1070, pci bus id: 0000:01:00.0)
有人可以向我解释吗？
编辑：好的，看来，我没有看到用法足够准确。GPU 是实际使用的，但只是在小高峰。大部分工作仍然由 CPU 完成。我正在运行一个具有 3 个卷积层和 2 个全连接层的 CNN。但这不可能吧？
] 1