[英]GPU control using CUDA in Python
I want to use Tensorflow-gpu in Python.我想在 Python 中使用 Tensorflow-gpu。 My PC has two GPUs.
我的电脑有两个 GPU。 (GPU-A and GPU-B)
(GPU-A 和 GPU-B)
I want to specify the GPU to be used.我想指定要使用的 GPU。
So when I run the program, I used the following command:所以当我运行程序时,我使用了以下命令:
CUDA_VISIBLE_DEVICES = 1
The GPU to be used can be specified according to the value.可以根据该值指定要使用的GPU。 Specifically, it was assigned as follows.
具体而言,分配如下。
CUDA_VISIBLE_DEVICES = 0
I was able to use GPU-B.我能够使用 GPU-B。
On the other hand另一方面
CUDA_VISIBLE_DEVICES = 1
At that time, I was able to use GPU-A.那时,我能够使用GPU-A。
However, I cannot know in advance whether GPU-A or GPU-B corresponds to the value 0 or 1. I found a statement that "this index is assigned in descending order of computing power".但是,我无法提前知道GPU-A或GPU-B对应的值是0还是1。我发现了“这个索引是按计算能力降序分配”的说法。 However, GPU-A can calculate faster than GPU-B.
但是,GPU-A 的计算速度比 GPU-B 快。 I think that "indexes are assigned in order of the revision number (Compute Capability), not the calculation speed".
我认为“索引是按照修订号(计算能力)而不是计算速度的顺序分配的”。 In fact, GPU-B has higher compute capability than GPU-A.
实际上,GPU-B 的计算能力比 GPU-A 更高。
If the hypothesis is correct, I have another question.如果假设是正确的,我还有一个问题。 If multiple GPUs with the same Compute Capability are used, how will they be assigned?
如果使用多个具有相同计算能力的 GPU,它们将如何分配?
You can try running nvidia-smi
in your terminal.您可以尝试在终端中运行
nvidia-smi
。 The output you get shows you GPU ordering.您获得的输出显示了 GPU 排序。 For example:
例如:
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 410.104 Driver Version: 410.104 CUDA Version: 10.0 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce GTX 108... Off | 00000000:01:00.0 Off | N/A |
| 0% 52C P8 21W / 250W | 2MiB / 11176MiB | 0% Default |
+-------------------------------+----------------------+----------------------+
| 1 GeForce GT 1030 Off | 00000000:02:00.0 On | N/A |
| 41% 54C P0 N/A / 30W | 1055MiB / 2001MiB | 6% Default |
+-------------------------------+----------------------+----------------------+
Here, I have 2 GPUs:在这里,我有 2 个 GPU:
You can then further use indexes in TensorFlow
.然后,您可以进一步在
TensorFlow
使用索引。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.