We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
http://unix.stackexchange.com/questions/38560/gpu-usage-monitoring-cuda
$ nvidia-smi