Intel® VTune™ Amplifier XE and VTune™ Performance Analyzer

I'm still confused by how to calculate the L1, L2 cache miss ratio after reading many related posts.

I'm trying to use Vtune to get the L1I, L1D, and L2 cache miss ratio on the platform of  Intel Xeon core microarchitecture.

First of all, the miss ratio I'm trying to get is the one under traditional definition like L2 misses number/whole L2 requests,not the one that defined in the Intel manual to calculate the ratio of L2 misses number of whole instruction retired like  L2_LINE_MISS.SELF.ANY/INST_RETIRED.ANY. 

Therefore, my question is:

1).When it comes to L1 cache miss ratio, I'm using the following formula by the meaning of the hardware events literally​:

How to find the microarchitecture code name? I want to analyse the performance of hard ware issues.

I try to use  VTune™ Amplifier XE 2011 to test the performance hardware issues of my C code project. 

But when I built a new amplifier project and attended to start a new Analysis for hardware Issues on type of "Advanced Intel(R) Microarchitecture Code Name Nehalem Analysis", it told me the processor should be based on the Intel(R) microarchitecture code name Nehalem and I couldn't continue.

There are no hardware events for LLC on the platform of Intel Xeon Core processor.

I want to get the miss rate of LLC on the platform of Core.

But while I use the cunstom analysis and try to pick the hardware events related to the LLC from the available events, I cannot find any.

Does that mean that vtune doesn't have any LLC related hardware events for Core processor?

Any help would be appreciated!

 

Sun.

Export results to CSV

I run an analysis in Amplifier XE 2013 GUI collecting hardware event counts.

Under the Top-down Tree tab, I am able to visualize the total event counts for each selected counter as a function of either the CPU timestamp or the elapsed application runtime. I would like to be able export this time-series data, rather than just the cumulative results, to a format such as CSV for analysis through an external program.

Profiling OpenMP* applications with Intel® VTune™ Amplifier XE

For those who are interested in profiling OpenMP* codes with Intel(R) VTune(TM) Amplifier XE, there is a new article describing all needed steps, tricks and results treatment with simple examples: 

http://software.intel.com/en-us/articles/profiling-openmp-applications-with-intel-vtune-amplifier-xe

Error using Vtune

I am trying to use Vtune and get the following error. There isnt any trace that would help me investigate this. Any pointers about how to investigate / fix this? Fwiw the error being returned appears to correspond to http://support.microsoft.com/kb/185294 and http://blogs.msdn.com/b/oldnewthing/archive/2010/07/30/10044061.aspx

Analysis started for C:\Windows\SysWOW64\cmd.exe (pid = 5732) 

Where can we find all the hardware events for a specific microarchitecture?

Recently,I've been using Vtune to get each cache level miss ratio on different platforms.

However, the hardware events that we used to calculate the miss ratio varies on different platform.

For instance, I have 3 types of machines: core , sandybridge and Ivybridge.

And some of the linux system can just use command line to start up the vtune.

Therefore , I'm wodering is there any commands of Vtune for users to check the hardware events that we can use on this machine?

Or are there any Intel mannual tell something about this?

Subscribe to Intel® VTune™ Amplifier XE and VTune™ Performance Analyzer