Is it normal that when I run "nvidia-smi" on Docker, it doesn't show up in the "Processes" section?
I'm building an environment for deep learning on ubuntu with Docker + GPU on ubuntu.
I think it's almost done, but there is one thing that bothers me. When I do "nvidia-smi" on Ubuntu, I see "processes".
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 440.100      Driver Version: 440.100      CUDA Version: 10.2     |
|-------------------------------+----------------------+----------------------+
| 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 166...  Off  | 00000000:01:00.0  On |                  N/A |
| 42%   37C    P8     8W / 125W |    249MiB /  5936MiB |      3%      Default |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID   Type   Process name                             Usage      |
|=============================================================================|
|    0      1127      G   /usr/lib/xorg/Xorg                            35MiB |
|    0      2006      G   /usr/lib/xorg/Xorg                            94MiB |
|    0      2202      G   /usr/bin/gnome-shell                          97MiB |
|    0      6565      G   /usr/lib/firefox/firefox                       2MiB |
|    0      7875      G   /usr/lib/firefox/firefox                       2MiB |
|    0     10070      G   /usr/lib/firefox/firefox                       2MiB |
+-----------------------------------------------------------------------------+
When I do 'nvidia-smi' on Docker, I don't see the 'processes'.
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 440.100      Driver Version: 440.100      CUDA Version: 10.2     |
|-------------------------------+----------------------+----------------------+
| 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 166...  Off  | 00000000:01:00.0  On |                  N/A |
| 42%   36C    P8     8W / 125W |    342MiB /  5936MiB |      2%      Default |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID   Type   Process name                             Usage      |
|=============================================================================|
+-----------------------------------------------------------------------------+
When I run "Jupyter Notebook", the GPU seems to be running.
”It's a spec," I read an article written a few years ago.
Is it a "spec" that Process is still not showing up today? Or is it because I'm not doing the right installation?
Lend me your wisdom. Thanks in advace!
Yes, you will not be able to see, due to driver not being aware of PID namespace. You can peruse the thread and the work-around using Python in particular, at
https://github.com/NVIDIA/nvidia-docker/issues/179#issuecomment-598059213
(I presume you are not using a VM, since persistence mode is OFF in the log shown).
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With