Vtune profiler- FP arithmetic is always zero.

[Surya_venkatesh]

Hi,

     I am new to Vtune profiler and i am not from computer science background. I was doing some basic Floating point operations like c=a*b  to understand Vtune results.  Both a and b are floats. I checked microarchitecture exploration and FP arithmetic was always shown 0.0%. I don't understand why it is zero considering what i do all is floating point arithmetics. Any info on that would be appreciated.

     I also have another question. I have two algorithms basically doing the same thing. I would like to compare the execution/run efficiency. Can i use instruction retired as main metric to compare them? Thank you very much in advance.

Regards,

Surya.

[RemyaP_Intel]

Hi,

Thank you for posting in Intel communities.

Could you please share with us a sample reproducer of the FP operation you were using to perform the microarchitecture exploration.

Regarding the second question:

You can perform Microarchitecture Exploration analysis to understand how efficiently your code is passing through the core pipeline. During Microarchitecture Exploration analysis, the VTune Profiler collects a complete list of events. It calculates a set of predefined ratios used for the metrics and facilitates identifying hardware-level performance problems.

For performance analysis, you may check how many instructions started their execution in the pipeline and compare the number with the count of retired operations. High difference shows that CPU does a lot of useless work and uses excess power.

You can read more here.https://www.intel.com/content/www/us/en/docs/vtune-profiler/user-guide/2023-0/general-exploration-analysis.html

We can assist you better, if you can provide the algorithms you are trying to compare.

--

Regards,

Remya Premdas

[RemyaP_Intel]

Hi,

We haven't heard back from you. Is your issue resolved?

--

Thanks,

Remya Premdas

[RemyaP_Intel]

Hi,

We assume that your issue is resolved. If you need any additional information, please post a new question as this thread will no longer be monitored by Intel.

--

Regards,

Remya Premdas