Get Started with Intel Arc GPU Monitoring: How to Show GPU Usage and Graphics Memory Usage

ai
,

How to Show Gpu Usage After Installing Intel-Gpu-Tools

After installing intel-gpu-tools, you can show GPU usage using the intel_gpu_top command.
Here are the steps to show GPU usage:

  1. Open a terminal window.
  2. Type the following command and press Enter to launch intel_gpu_top: intel_gpu_top
  3. The intel_gpu_top tool will display real-time GPU usage statistics for the Intel GPU. You will see a screen with a list of metrics that shows the utilization of various GPU components, such as the GPU engine, memory, and the GPU core frequency.
  4. Press ‘q’ to exit intel_gpu_top.

GPU
The result of this command would be like the image above.

Note: Make sure you have the appropriate drivers installed and your GPU is supported by intel-gpu-tools.

What Else Are Included in the Intel-Gpu-Tools

intel-gpu-tools is a collection of tools for debugging and monitoring Intel GPUs on Linux. In addition to intel_gpu_top, which shows real-time GPU usage statistics, intel-gpu-tools includes several other useful tools, including:

  1. intel_gpu_dump: This tool can be used to extract debug information from the Intel GPU, including error state and performance counters.
  2. intel_reg_read/intel_reg_write: These tools can be used to read or write registers in the Intel GPU, which can be useful for debugging and testing.
  3. intel_audio_dump: This tool can be used to capture audio playback data from Intel GPUs with integrated audio controllers.
  4. intel_gpu_frequency: This tool can be used to monitor or set the frequency of the Intel GPU.
  5. intel_display_power: This tool can be used to control the display power of the Intel GPU, which can be useful for power management and reducing energy consumption.

Overall, intel-gpu-tools is a useful collection of tools for developers and system administrators who need to monitor or debug Intel GPUs on Linux systems.

How to Show Graphics Memory Usage

Sysmon can be utilized to monitor GPU memory, but the tool is not currently available in a system package. Therefore, it must be compiled from the source code.

1
2
3
4
git clone https://github.com/intel/pti-gpu/
cd pti-gpu/tools/sysmon
mkdir build && cd build
cmake -DCMAKE_BUILD_TYPE=Release .. && make

Once the tool has been compiled, it can be executed via the command ./sysmon.

graphics memory

The graphics memory usage of the Arc card is displayed in the Device Memory Used(MB) column, which is partitioned by process IDs.

The compiled file can be copied into the /bin/ directory or any other directory within the $PATH environment variable. Once the file is in the correct directory, you can run the tool using the sysmon command.