Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I have the cpu-usage profile of some standard benchmarks in which I can see the cpu-usage to be varying a lot . I wish to increase the cpu-frequency to the highest available value , exactly at the instants the cpu starts executing compute-intensive portions of the workload and reduce the cpu-frequency at instants when workload is not compute-intensive. I'm using linux OS, on a cluster of 2 core2duo machines

Please let me know what code I'm supposed to modify ? Should I change the P-state governor in order to do this ?

share|improve this question

Just enable the "ondemand" cpufreq governor on your Linux distro. No need to modify any code.

On Redhat/Fedora, just install the cpuspeed package and and chkconfig cpuspeed on if it hasn't been done already by default.

share|improve this answer
Personally, I prefer 'conservative' because 'ondemand' cpu speeds tend to swing widely up and down while 'conservative' moves less. – sybreon Jan 16 '10 at 5:14

Most decent modern servers will manage this themselves - you haven't told us anything about your machine, processors, OS etc.

share|improve this answer
I'm using a cluster of 2 nodes running linux on core2duo processors – kashyapa Jan 15 '10 at 23:41

Read this:

TL;DR: you're better off running powertop and making sure your box can properly go idle(C states)

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.