Performance Monitor Next

A lightweight, real-time system performance monitor for Windows, built with Python. It provides hardware and runtime metrics directly in your terminal or through an HTTP API.

Features

• CPU monitoring: temperature, clock, usage, load, voltage, and power.
• GPU monitoring:
  • General GPU metrics via LibreHardwareMonitor.
  • NVIDIA GPU metrics via NVML.
• Memory monitoring: physical memory and swap usage.
• Network monitoring: real-time network activity.
• Frame-time monitoring: rendering/frame-time insights.
• Live updates: configurable refresh interval.
• CLI + Server modes: terminal dashboard or HTTP endpoint.

Requirements

• Operating system: Windows (depends on LibreHardwareMonitor and pythonnet).
• Python: 3.11 or higher.
• Permissions: run as Administrator to access all hardware sensors.

Installation

git clone https://github.com/LoHhhha/performance_monitor_next.git
cd performance_monitor_next
pip install .

Quick Start

Run as Terminal Dashboard

python -m performance_monitor.cmd.runner

Useful arguments:

• -ft, --flush_time: refresh interval in seconds (default: 0.8).
• --exclude-general-gpu: disable general GPU monitoring.
• --exclude-nvidia-gpu: disable NVIDIA GPU monitoring.

Example:

python -m performance_monitor.cmd.runner -ft 1.0

Run as HTTP Server

python -m performance_monitor.server.runner

Useful arguments:

• -p, --port: server port (default: 54321).
• --host: host address (default: 127.0.0.1).
• --path: metrics path (default: /info).

Example (serve at http://127.0.0.1:8000/info):

python -m performance_monitor.server.runner --host 127.0.0.1 -p 8000 --path /info

Screenshots

The layout adapts automatically to terminal width.

Wide terminal view · Narrow terminal view

Troubleshooting

• No sensor data / missing values: run the terminal as Administrator.
• GPU fields are empty: ensure GPU drivers are installed and supported.
• pythonnet/hardware access issues: verify Python version and reinstall dependencies.

Notes

• This project is Windows-focused by design due to hardware dependency constraints.