gpulink 0.2.0

A simple tool for monitoring and displaying GPU stats

gpulink

A library for monitoring and displaying NVIDIA GPU stats.
gpulink uses pynvml - a Python wrapper for the NVIDIA Management Library (NVML).

Current status

⚠️!!! This project is under heavy development !!!⚠️

Installation

Installation using PIP

To install gpulink using the Python Package Manager (PIP) run:
pip install gpulink

gpulink can also be used from source. Here perform the following steps to create a Python environment and to install the requirements:

1. Create an environment: python -m venv env
2. Activate the environment: .\env\Scripts\Activate
3. Install requirements: pip install -r requirements.txt

Usage

gpulink can be used as a library as well as from the command line.

Library usage

To integrate gpulink to a Python script, import gpulink and create an NVContext. This context manages the creation and destruction of the nvml session and provides several query and utility functions ( see API example):

import gpulink as gpu

with gpu.NVContext() as ctx:
   print(f"Available GPUs: {ctx.gpu_names}")
   memory_information = ctx.get_memory_info(ctx.gpus)
   ...

gpulink provides a Recorder class for recording several GPU properties. An instance of this class mst be created using one of the factory methods, e.g.:

    recorder = gpu.Recorder.create_memory_recorder(ctx, ctx.gpus)
    recorder.start()
    ... # Do some GPU stuff
    recorder.stop(auto_join=True)

One a recording is finished, the data can be accessed:

recording = recording = recorder.get_recording()

gpulink provides a Plot class for visualizing recordings using matplotlib:

    from pathlib import Path
    
    plot = gpu.Plot(recording)
    plot.plot(scale_y_axis=True)
    plot.save(Path("memory.png"), scale_y_axis=True)
    
    figure, axis = plot.generate_graph()  # The generated Figure and Axis can also be accessed directly.

Command-line usage

During installation, gpulink also registers a command-line script accessible through the gpulink command.

usage: gpulink [-h] {sensors,record} ...

GPU-Link: Monitor NVIDIA GPU status

positional arguments:
  {sensors,record}

optional arguments:
  -h, --help        show this help message and exit

Examples

• View GPU sensor status: gpulink sensors

╒════════╤═════════════════════╤═════════════╤═════════════════╤═══════════════╕
│ GPU    │ Memory [MB]         │   Temp [°C] │   Fan speed [%] │ Clock [MHz]   │
╞════════╪═════════════════════╪═════════════╪═════════════════╪═══════════════╡
│ GPU[0] │ 1588 / 25769 (6.2%) │          34 │              41 │ Graph.: 173   │
│        │                     │             │                 │ Memory: 403   │
│        │                     │             │                 │ SM: 173       │
│        │                     │             │                 │ Video: 539    │
╘════════╧═════════════════════╧═════════════╧═════════════════╧═══════════════╛

• Record gpu memory information and save a plot as PNG: gpulink record -o memory.png

View GPU sensor status

Command:

gpulink sensors

Output:

╒═════════════════════╤═════════════════╕
│ Record duration [s] │ Frame rate [Hz] │
├─────────────────────┼─────────────────┤
│ 14.0294178          │ 235             │
╘═════════════════════╧═════════════════╛
╒═════╤══════════════════╤══════════════════════╕
│ GPU │ Name             │ Memory used [MB]     │
├─────┼──────────────────┼──────────────────────┤
│ 0   │ NVIDIA TITAN RTX │ minimum: 1584.754688 │
│     │                  │ maximum: 2204.585984 │
╘═════╧══════════════════╧══════════════════════╛