gpustat =========

[][pypi_gpustat]

Just less than nvidia-smi?

NOTE: This works with NVIDIA Graphics Devices only, no AMD support as of now. Contributions are welcome!

Self-Promotion: A web interface of gpustat is available (in alpha)! Check out [gpustat-web][gpustat-web].

[gpustat-web]: https://github.com/wookayin/gpustat-web

Quick Installation

---

Install from [PyPI][pypi_gpustat]:

pip install gpustat

If you don't have root (sudo) privilege, please try installing gpustat on user namespace: pip install --user gpustat.

To install the latest version (master branch) via pip:

pip install git+https://github.com/wookayin/gpustat.git@master

NVIDIA Driver and pynvml Requirements

>[!IMPORTANT]

DO NOT: pip install pynvml, nor include [pynvml][pypi_wrong] as a dependency in your python project. This will not work.

>

Instead: pip install nvidia-ml-py. [nvidia-ml-py][pypi_pynvml] is NVIDIA's the official python binding for NVML.

• gpustat 1.2+: Requires nvidia-ml-py >= 12.535.108 ([#161][gh-issue-161])
• gpustat 1.0+: Requires NVIDIA Driver 450.00 or higher and nvidia-ml-py >= 11.450.129.
• If your NVIDIA driver is too old, you can use older gpustat versions (pip install gpustat<1.0). See [#107][gh-issue-107] for more details.

Python requirements

• gpustat<1.0: Compatible with python 2.7 and >=3.4
• gpustat 1.0: [Python >= 3.4][gh-issue-66]
• gpustat 1.1: Python >= 3.6

Usage

---

$ gpustat

Options (Please see gpustat --help for more details):

• --color : Force colored output (even when stdout is not a tty)
• --no-color : Suppress colored output
• -u, --show-user : Display username of the process owner
• -c, --show-cmd : Display the process name
• -f, --show-full-cmd : Display full command and cpu stats of running process
• -p, --show-pid : Display PID of the process
• -F, --show-fan : Display GPU fan speed
• -e, --show-codec : Display encoder and/or decoder utilization
• -P, --show-power : Display GPU power usage and/or limit (draw or draw,limit)
• -a, --show-all : Display all gpu properties above
• --id : Target and query specific GPUs only with the specified indices (e.g. --id 0,1,2)
• --no-processes : Do not display process information (user, memory) ([#133][gh-issue-133])
• --watch, -i, --interval : Run in watch mode (equivalent to watch gpustat) if given. Denotes interval between updates.
• --json : JSON Output ([#10][gh-issue-10])
• --print-completion (bash|zsh|tcsh) : Print a shell completion script. See [#131][gh-issue-131] for usage.

Tips

• Try gpustat --debug if something goes wrong.
• To periodically watch, try gpustat --watch or gpustat -i ([#41][gh-issue-41]).

- For older versions, one may use watch --color -n1.0 gpustat --color.

• Running nvidia-smi daemon (root privilege required) will make querying GPUs much faster and use less CPU ([#54][gh-issue-54]).
• The GPU ID (index) shown by gpustat (and nvidia-smi) is PCI BUS ID,

while CUDA uses a different ordering (assigns the fastest GPU with the lowest ID) by default. Therefore, in order to ensure CUDA and gpustat use same GPU index, configure the CUDADEVICEORDER environment variable to PCIBUSID (before setting CUDAVISIBLEDEVICES for your CUDA program): export CUDADEVICEORDER=PCIBUSID.

Python API

gpustat can also be used as a library to query GPUs from your own code:

import gpustat
stats = gpustat.new_query()   # a GPUStatCollection stats.print_formatted()       # same output as the gpustat command
data = stats.jsonify()        # query result as a dict (same as gpustat --json)

[pypi_gpustat]: https://pypi.org/project/gpustat/ [pypi_pynvml]: https://pypi.org/project/nvidia-ml-py/#history [pypi_wrong]: https://pypi.org/project/pynvml/ [gh-issue-10]: https://github.com/wookayin/gpustat/issues/10 [gh-issue-41]: https://github.com/wookayin/gpustat/issues/41 [gh-issue-54]: https://github.com/wookayin/gpustat/issues/54 [gh-issue-66]: https://github.com/wookayin/gpustat/issues/66 [gh-issue-107]: https://github.com/wookayin/gpustat/issues/107 [gh-issue-131]: https://github.com/wookayin/gpustat/issues/131 [gh-issue-133]: https://github.com/wookayin/gpustat/issues/133 [gh-issue-161]: https://github.com/wookayin/gpustat/issues/161#issuecomment-1784007533

Default display

---

[0] GeForce GTX Titan X | 77°C,  96 % | 11848 / 12287 MB | python/52046(11821M)

• [0]: GPU index (starts from 0) as PCIBUSID
• GeForce GTX Titan X: GPU name
• 77°C: GPU Temperature (in Celsius)
• 96 %: GPU Utilization
• 11848 / 12287 MB: GPU Memory Usage (Used / Total)
• python/...: Running processes on GPU, owner/cmdline/PID (and their GPU memory usage)

Changelog

---

See CHANGELOG.md

License

---

MIT License