ace-gpu-1_installation_log

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision		 Previous revision
ace-gpu-1_installation_log [2017/03/24 20:24] – csteelace-gpu-1_installation_log [2024/03/26 13:52] (current) – external edit 127.0.0.1
Line 30: Line 30:
  
 <code> <code>
-01:00.0 VGA compatible controller: NVIDIA Corporation Device 1b00 (rev a1)+01:00.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1)
 01:00.1 Audio device: NVIDIA Corporation Device 10ef (rev a1) 01:00.1 Audio device: NVIDIA Corporation Device 10ef (rev a1)
 +02:00.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1)
 +02:00.1 Audio device: NVIDIA Corporation Device 10ef (rev a1)
 +03:00.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1)
 +03:00.1 Audio device: NVIDIA Corporation Device 10ef (rev a1)
 +
 </code> </code>
  
Line 121: Line 126:
  
 <code> <code>
-In process+# set user PATH to include /usr/local/cuda-8.0/bin 
 +if [ -d "/usr/local/cuda-8.0/bin" ]; then 
 +    PATH="/usr/local/cuda-8.0/bin:$PATH" 
 +fi 
 + 
 +# set user LD_LIBRARY_PATH to include /usr/local/cuda-8.0/lib64 
 +if [ -d "/usr/local/cuda-8.0/lib64" ]; then 
 +    LD_LIBRARY_PATH="/usr/local/cuda-8.0/lib64:$LD_LIBRARY_PATH" 
 + 
 + 
 +</code> 
 +<code> 
 +</code> 
 +<code>
 </code> </code>
 +
 +=== GPU Accounting Setup and Configuration ===
 +
 +The CUDA nvidia-persistenced needs to be configured for the target OS's startup system. In the case of Ubuntu 16.04 this would be systemd. 
 +
 +== Confirm Driver Version ==
  
 <code> <code>
 +nvidia-smi
 </code> </code>
 +
 +Output example
 +
 <code> <code>
 +Wed Apr 26 14:20:40 2017       
 ++-----------------------------------------------------------------------------+
 +| NVIDIA-SMI 375.39                 Driver Version: 375.39                    |
 +|-------------------------------+----------------------+----------------------+
 +| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
 +| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
 +|===============================+======================+======================|
 +|   0  TITAN X (Pascal)    Off  | 0000:01:00.0      On |                  N/A |
 +| 41%   68C    P2    95W / 250W |   1902MiB / 12186MiB |     96%      Default |
 ++-------------------------------+----------------------+----------------------+
 +                                                                               
 ++-----------------------------------------------------------------------------+
 +| Processes:                                                       GPU Memory |
 +|  GPU       PID  Type  Process name                               Usage      |
 +|=============================================================================|
 +|    0      1977    G   /usr/lib/xorg/Xorg                              60MiB |
 +|    0     10406    C   /data1/data/kwagstyl/anaconda2/bin/python      189MiB |
 +|    0     10660    C   /data1/data/kwagstyl/anaconda2/bin/python     1495MiB |
 +|    0     16118    C   ...freesurfer_LBL/bin/mris_fix_topology_cuda   153MiB |
 ++-----------------------------------------------------------------------------+
 </code> </code>
 +
 +== Download and uncompress ==
 +
 +Download nvidia-persistenced version that matches your driver version (see output example above)
 +
 <code> <code>
 +mkdir -p ~/src/ubuntu/16.04/nvidia
 +cd ~/src/ubuntu/16.04/nvidia
 +wget ftp://download.nvidia.com/XFree86/nvidia-persistenced/nvidia-persistenced-375.39.tar.bz2
 +tar xvjf nvidia-persistenced-375.39.tar.bz2
 </code> </code>
 +
 +== Edit the nvidia-persistenced.conf.template ==
 +
 += Confirm creation of the nvidia-persistenced user = 
 +
 <code> <code>
 +sudo cat /etc/passwd | grep nvidia
 +</code>
 +
 +Output example:
 +
 +<code>
 +nvidia-persistenced:x:126:132:NVIDIA Persistence Daemon,,,:/:/sbin/nologin
 +</code>
 +
 +== Edit the systemd template ==
 +
 +<code>
 +cd nvidia-persistenced-375.39/init/systemd
 +nano nvidia-persistenced.service.template
 +</code>
 +
 +Replace __USER__ with the nvidia-persistence users name `nvidia-persistenced`
 +
 +== Run the installer ==
 +
 +<code>
 +cd ~/sys/sw/ubuntu/16.04/nvidia/nvidia-persistenced-375.39/init
 +sudo ./install.sh
 +</code>
 +
 += Output example =
 +
 +<code>
 +Checking for common requirements...
 +  sed found in PATH?  Yes
 +  useradd found in PATH?  Yes
 +  userdel found in PATH?  Yes
 +  id found in PATH?  Yes
 +Common installation/uninstallation supported
 +
 +Creating sample System V script... done.
 +Creating sample systemd service file... done.
 +Creating sample Upstart service file... done.
 +
 +Checking for systemd requirements...
 +  /usr/lib/systemd/system directory exists?  No
 +  /etc/systemd/system directory exists?  Yes
 +  systemctl found in PATH?  Yes
 +systemd installation/uninstallation supported
 +
 +Installation parameters:
 +  User  : nvidia-persistenced
 +  Group : nvidia-persistenced
 +  systemd service installation path : /etc/systemd/system
 +
 +User 'nvidia-persistenced' already exists, skipping useradd...
 +User 'nvidia-persistenced' is in primary group 'nvidia-persistenced'.
 +Stopping nvidia-persistenced.service... done.
 +Installing sample systemd service nvidia-persistenced.service... done.
 +Enabling nvidia-persistenced.service... done.
 +Starting nvidia-persistenced.service... done.
 +</code>
 +
 +== Check ==
 +
 +<code>
 +sudo service nvidia-persistenced status
 +</code>
 +
 +== Troubleshooting ==
 +
 +add "--persistence-mode --verbose" to the line where the service is started.
  
  • ace-gpu-1_installation_log.1490387088.txt.gz
  • Last modified: 2024/03/26 13:52
  • (external edit)