怎么查看进程消耗GPU

使用nvidia-smi可以查看到进程消耗的GPU情况
但是无法显示完整进程GPU

1. 使用pid查询
   使用nvidia-smi命令查看消耗的GPU已经进程号

# nvidia-smi 
Thu Jun 25 11:11:13 2026       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 580.159.03             Driver Version: 580.159.03     CUDA Version: 13.0     |
+-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 4090        Off |   00000000:18:00.0 Off |                  Off |
| 30%   25C    P8             28W /  450W |   12980MiB /  24564MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   1  NVIDIA GeForce RTX 4090        Off |   00000000:3B:00.0 Off |                  Off |
| 30%   31C    P5             69W /  450W |    2110MiB /  24564MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   2  NVIDIA GeForce RTX 4090        Off |   00000000:86:00.0 Off |                  Off |
| 30%   25C    P8             22W /  450W |      18MiB /  24564MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI              PID   Type   Process name                        GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|    0   N/A  N/A            3512      G   /usr/lib/xorg/Xorg                        4MiB |
|    0   N/A  N/A           27256      C   python                                  906MiB |
|    0   N/A  N/A           42270      C   python                                  444MiB |
|    0   N/A  N/A           42271      C   python                                  538MiB |
|    0   N/A  N/A           42357      C   python                                 2108MiB |
|    0   N/A  N/A           43393      C   python                                  436MiB |
|    0   N/A  N/A           43397      C   python                                  538MiB |
|    0   N/A  N/A           43520      C   python                                 2098MiB |
|    0   N/A  N/A           61464      C   python                                 1916MiB |
|    0   N/A  N/A           66625      C   python                                  906MiB |
|    0   N/A  N/A           73081      G   ...m_detection/.venv/bin/python3         20MiB |
|    0   N/A  N/A          101530      C   python                                  902MiB |
|    0   N/A  N/A         1875220    C+G   ...m_detection/.venv/bin/python3       2014MiB |
|    1   N/A  N/A            3512      G   /usr/lib/xorg/Xorg                        4MiB |
|    1   N/A  N/A           73081      C   ...m_detection/.venv/bin/python3       2086MiB |
|    2   N/A  N/A            3512      G   /usr/lib/xorg/Xorg                        4MiB |
+-----------------------------------------------------------------------------------------+

这里列出的进程名只有python没有完整的进程名称不知道是那个进程消耗的
根据pid查询

# ps xua|grep 27256| grep -v grep
root       27256  0.0  0.3 26302428 971624 pts/4 Sl+  5月22  35:39 python infer.py

2. 使用nvtop命令查询
   上面的命令操作需要多一个步骤
   下面使用nvtop命令查看
   安装nvtop

# apt install nvtop

查看

# nvtop

底部按GPU从高到低进行了排序并且列出了完整的进程路径