nvitop is an interactive NVIDIA device and process monitoring tool. It has a colorful and informative interface that constantly updates the status of devices and processes. As a resource monitor, it includes many features and options, such as tree view, environment variable viewing, process filtering, process indicator monitoring, etc.Among other things, the package provides deep learning researchers with a CUDA device selection toolnvisel. It also provides a convenient API that allows developers to write their own monitoring tools.


  • Informative and beautiful output: display rationvidia-smiColorful beautiful box drawing for more information.
  • monitor mode: can be run as a resource monitor instead of just printing the result once.

    • Bar and History Graphs
    • process ordering
    • process filter
    • Send a signal to a process with a keystroke
    • Tree view screen of a GPU process and its parent process
    • Environment variables screen
    • help screen
    • mouse support

  • interactive: Responds to user input (from keyboard and/or mouse) while in monitor mode. (Comparedgpustatandpy3nvml)
  • efficient:

    • Use the NVML Python bindings directlyQuery device status, not resolvenvidia-smioutput (withnvidia-htopcompared to)
    • Support for sparse queries and cache results using TTLCache from cachetools (withgpustatcompared to)
    • Use the curses library to display information instead of printing it with ANSI escape codes (as withpy3nvmlcompared to)
    • Use multithreading to collect information asynchronously and respond to user input more quickly. (andnvtopcompared to)

  • portable: Available for Linux and Windows.

    • Use cross-platform libraries psutil Get the host process information instead of calling it in the child processps -p <pid>. (andnvidia-htopandpy3nvmlcompared to)
    • Written in pure Python, easy to usepipInstall (withnvtopcompared to)

  • Can be integrated: Easy to integrate into other applications, not just monitoring. (Comparednvidia-htopandnvtop)

#nvitop #Homepage #Documentation #Downloads #OneStop #Solution #GPU #Process #Management #News Fast Delivery

Leave a Comment

Your email address will not be published. Required fields are marked *