DOS GPU Usage

From Tuflow
Jump to navigation Jump to search

Nvidia-smi (also NVSMI) provides monitoring and management capabilities for each of NVIDIA's Tesla, Quadro, GRID and GeForce GPU device from Fermi and higher architecture families. The following steps can be used to access nVidia-smi and review real-time GPU usage statistics for TUFLOW simulations:

Accessing nvidia-smi to review GPU Usage

  1. Launch the DOS Command Prompt from the Run window (press Win+R on your keyboard to open "run" then type cmd).
    Nvidia-smi 001.png
  2. Change the directory location to the folder where nvidia-smi is located. Type cd C:\Program Files\NVIDIA Corporation\NVSMI into the DOS window and press enter.
    Nvidia-smi 002.png
  3. Type nvidia-smi -l 10 in the DOS window and press enter. This will instruct nvidia-smi to refresh every 10 seconds.
    Nvidia-smi 003.png
  4. Review the nvidia-smi usage summary.
    Nvidia-smi 004.png

What to do if GPU utilisation = 0% when running TUFLOW HPC (when using the GPU Module)

A common cause of 0% utilisation is if the calculations have accidentally been paused by activating 'Quick Edit' in the console. Quick edit mode was introduced in Windows 10 (in Windows not TUFLOW) and is initiated if the curser clicks somewhere on the DOS window while a TUFLOW simulation is running. Quick Edit mode can be deactivated to avoid this issue.

  1. Right click the DOS window header. Select Properties.
    Nvidia-smi 005.png
  2. Uncheck Quick Edit Mode. This will turn it off for the current session.
    Nvidia-smi 006.png
  3. Update the default DOS window properties so this becomes the default mode. Right click the DOS window header. Select Defaults.
    Nvidia-smi 007.png
  4. Uncheck Quick Edit Mode, just like in the "Properties" dialog box . This will turn Quick Edit Mode off for all future sessions.

Other troubleshooting tips:

  • Ensure the simulation has entered the calculation loop and is not currently in the initialisation phase (which is performed on the CPU and not GPU)
  • Check TUFLOW is running on a GPU device by checking the console title includes the GPU Devices text. If this text is not included, then the command "Hardware == GPU" was most likely not found in the TCF
GPU devices console.PNG