I am trying to figure out for different workloads (CPU intensive, high DB IOPs etc), which compute series of cloud VM would be ideal.

In order to do that I need to be able to get the CPU utilisation. I am using python psutil library to get CPU % (similar to top) , how can I determine actual CPU usage given number of cores of CPU.

Trying to find this I am confused by a lot of other aspects also. My main aim is to find which series [compute intensive, memory intensive, general-purpose etc] would be better for a given job. To find that what are the parameters that could affect that I need to consider. Are there any utilities, tools that would help in this case. I have used top and perf to check the CPU usage, but to run it programatically is there any standard way.

1 Answer 1


That's mostly vendor specific (Linux GPU monitoring has no good APIs) but there's a universal utility for all three (AMD, Intel and NVIDIA) which unfortunately is only good for monitoring but not for parsing:


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .