Software Tuning, Performance Optimization & Platform Monitoring
Discussion regarding monitoring and software tuning methodologies, Performance Monitoring Unit (PMU) of Intel microprocessors, and platform updating.

Slower code execution if compiled with Intel Parallel studio

Roman_I_
Beginner
454 Views

Hi, everyone,

I've just got intel parallel studio 2015 compiler and I'm trying to compile the code, that I previously compiled using Apple LLVM 6.0 compiler under xCode 6.0. I did not do any changes to build options except that I changed the compiler. Surprisingly execution time has increased 2.5 times after that. 

I'm having macbook pro with intel i7 running OSX 10.10. The main part of the program is working with vector data type.

If I switch back to LLVM compiler, the execution time is still 2.5 times longer and to reach the initial execution time I have to create whole project and copy .cpp files there.

What may be the reason of bad performance of intel compiler compared to LLVM, which seem to contaminate the LLVM?

0 Kudos
5 Replies
TimP
Honored Contributor III
454 Views

This case certainly looks as if performance would be data dependent (and depend on compile options), but you don't provide any data.   Did you look at the helpful optimization diagnostics offered by your compilers, with reference to where time is spent (which you don't give us the ability to determine)?

If you want to eliminate the possibility that memory leaks affect one of your builds, you may need to reboot.

0 Kudos
Roman_I_
Beginner
454 Views

I figured out that if I turn on optimization in LLVM compiler, it reduces time needed to compile 5 times and it cures the slow performance after using Intel compiler. I will try to find the optimization techniques for Intel compiler and hopefully it will outperform apple compiler.

My data is too huge (and it is proprietary) to be uploaded here, sorry :(

Thanks for the answer

0 Kudos
Bernard
Valued Contributor I
454 Views

Which ICC switches were used to compile your code?

0 Kudos
Roman_I_
Beginner
454 Views

I used -O3 -ip, but i could use these options only is I compiled from the terminal, I could not find these optimization options in xCode. with -O3 -ip the executions of icpc compiler program took only 30% more, than the same program compiled using LLVM

0 Kudos
Bernard
Valued Contributor I
454 Views

Did you inspect assembly code generated by both compilers?

0 Kudos
Reply