linux-hwinfo64 0.1.0

Add your description here

Linux HWINFO64

Written in Python
This is a scaled down proof of concept. It's a simple script to output a system's hardware information and some simple metrics for the CPU and GPU In the future, I might expand this to include more of the features of hwinfo64, but it's a simple MVP for now

Requirements

• nvidia-smi for Nvidia GPUs
• rocm-smi for AMD GPUs

GPU Detection

It first checks for NVIDIA GPUs using nvidia-smi
Then it checks for AMD GPUs by:

• Looking at vendor ID in /sys/class/drm/card0/device/vendor
• Using lspci to search for AMD graphics adapters
• The most accurate way to get usage is using rocm-smi, otherwise it relies on /sys/class/drm/card0/device/gpu_busy_percent[1]

Usage Notes:

CLI Arguments

• --graph or -g: Run the monitor in graph mode
• --record or -r: Record metrics to a CSV file for later analysis
• --output or -o: Specify the output CSV file name (default: hw_metrics.csv)

Graph Mode

A new display mode that shows line graphs of your hardware metrics over the last 2 minutes

• CPU usage history
• Memory usage history
• GPU utilization history (if available)
• GPU memory usage history (if available)

Notes

For AMD GPUs, some metrics might not be available depending on your specific card and drivers:

• Temperature reading paths can vary between different AMD cards
• Memory usage requires ROCm tools to be installed
• GPU utilization might not be available on older cards/drivers

Additional requirements for AMD GPU support:

• For basic detection: standard Linux utilities like lspci
• For more detailed metrics: AMD's ROCm tools (rocm-smi)

Links:

• Docs for installing rocm for AMD
  • https://rocm.docs.amd.com/projects/install-on-linux/en/latest/install/quick-start.html

[1] I don't know a lot about how linux determines device number, but it will increment the value for card0 if you ever change your graphics card (presumably it stores the previous device values / configs / whatever). I plan to investigate this further