We recommend you to submit your post to a Linux community or in our Intel® Software network for further assistance on this matter.
You may want to access the following link:
Also, we like to point out the fact that comparing these two processors may not be an equal comparison due to the fact that both CPU’s have different architectures and feature sets one is an Ivy bridge processor and the other is a Bloomfield.
Thanks for your reply.
For us, it is not a Linux issue. we have just our test that we tried on different Intel processors and always had good results.
for Xeon E3-1220v2 and i5 processor, we got such strange behaviors. we think that it is more management interrupts in the processor itself, or something related to pipelines or L0, because executing the same code 45000 starts to generate some interrupts which are seen in our measuremnts, while executing the code 5000, we do not see the interrupts anymore.
I don’t think we are able to help you from our normal Customer Support forum.
We were suggesting the Intel Software Forum or a Linux forum since they are generally more knowledgeable on how processors generate interrupts.
We were not implying that this was a Linux issue. Sorry for any misunderstanding.