Hi, I recently looked for one and couldn't find exactly that but the closest seemed to be either dcgm: https://docs.nvidia.com/datacenter/dcgm/latest/user-guide/index.html or if you fancy writing your own tool to you can probe total power consumption at a given time from libnvml: https://docs.nvidia.com/deploy/nvml-api/group__nvmlDeviceQueries.html#group__nvmlDeviceQueries_1g732ab899b5bd18ac4bfb93c02de4900a
Best, Nicolas Le mardi 07 juin 2022 à 16:21 +0200, Matthias Leopold a écrit : Hi, I know this might be a too simple question for a bigger topic, but I'll just try: is there something like seff for measuring the efficiency of NVIDIA GPU usage in Slurm jobs? thx Matthias
