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.
Usage
$ gpustat
Options:
--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-P
,--show-power
: Display GPU power usage and/or limit (draw
ordraw,limit
)-a
,--show-all
: Display all gpu properties above--watch
,-i
,--interval
: Run in watch mode (equivalent towatch gpustat
) if given. Denotes interval between updates. (#41)--json
: JSON Output (Experimental, #10)
Tips
- To periodically watch, try
gpustat --watch
orgpustat -i
(#41).- For older versions, one may use
watch --color -n1.0 gpustat --color
.
- For older versions, one may use
- Running
nvidia-smi daemon
(root privilege required) will make the query much faster and use less CPU (#54). - The GPU ID (index) shown by
gpustat
(andnvidia-smi
) is PCI BUS ID, while CUDA differently assigns the fastest GPU with the lowest ID by default. Therefore, in order to make CUDA andgpustat
use same GPU index, configure theCUDA_DEVICE_ORDER
environment variable toPCI_BUS_ID
(before settingCUDA_VISIBLE_DEVICES
for your CUDA program):export CUDA_DEVICE_ORDER=PCI_BUS_ID
.
Quick Installation
Install from PyPI:
pip install gpustat
If you don't have root privilege, please try to install 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
Note that starting from v1.0, gpustat will support only Python 3.4+. For older versions (python 2.7, <3.4), you can continue using gpustat v0.x.
Changelog
See CHANGELOG.md