I want to see if the current workload is 100% using compute engine, in other words, how many hardware resource are used for the current task at the current time.
Is “nvidia-smi dmon -s u” what I should use?
I want to see if the current workload is 100% using compute engine, in other words, how many hardware resource are used for the current task at the current time.
Is “nvidia-smi dmon -s u” what I should use?
nvidia-smi
will show some statistics