---

nvsmifs

A user-friendly wrapper for nvidia-smi, forked from pmav99's repository, with added support for fan speed and power draw monitoring. This tool is particularly useful in multi-GPU systems, allowing you to filter GPUs based on resource usage (e.g., selecting the least utilized GPU).

---

Features

• Enhanced GPU Monitoring: Includes fan speed and power draw metrics.
• Resource-Based Filtering: Easily identify the least utilized GPU in your system.
• Flexible Usage: Use as a CLI tool or integrate it as a Python library.

---

Table of Contents

1. Usage
   • CLI
   • Library
2. Quick Example: GPU Info
3. Prerequisites
4. Installation
   • Using pipx (Recommended)
   • Using pip
5. Contributing
6. License

---

Usage

CLI

Run the following commands to explore available options:

nvsmifs --help
nvsmifs ls --help
nvsmifs ps --help

Library

You can also use nvsmifs as a Python library for more advanced workflows:

import nvsmifs

gpus = nvsmifs.get_gpus()
available_gpus = nvsmifs.get_available_gpus()
gpu_processes = nvsmifs.get_gpu_processes()

---

Quick Example: GPU Info

Here's a simple Python function to display detailed GPU information:

import nvsmifs

def gpu_info():
    for gpu in nvsmifs.get_gpus():
        if gpu.display_active == 'Enabled':
            gpu_details = (
                f"ID: {gpu.id}, Power: {gpu.power_draw}W, "
                f"Name: {gpu.name}, Temp: {gpu.temperature}°C, "
                f"Fan Speed: {gpu.fan_speed}%, Utilization: {gpu.gpu_util}%"
            )
            print(gpu_details)

gpu_info()

Output example:

ID: 0, Power: 50W, Name: NVIDIA GTX 1080, Temp: 65°C, Fan Speed: 30%, Utilization: 20%

---

Prerequisites

Ensure the following requirements are met before using nvsmifs:

• An NVIDIA GPU
• nvidia-smi installed and accessible
• Python 2.7 or 3.6+

---

Installation

Using pipx (Recommended)

The recommended installation method is via pipx. Install nvsmifs with:

pipx install nvsmifs

This creates a virtual environment in ~/.local/pipx/venvs/nvsmifs and adds the nvsmifs executable to ~/.local/bin.

Using pip

Alternatively, install directly with pip:

pip install --user nvsmifs

---

Contributing

Contributions are welcome! If you'd like to improve this project, please:

1. Fork the repository.
2. Create a feature branch.
3. Submit a pull request with a clear description of your changes.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---