Hello,
On Wed, Jul 16, 2008 at 3:52 PM, Juan Ángel Lorenzo <[EMAIL PROTECTED]> wrote: > Hi Stephane, > > We are considering using perfmon2 to estimate the performance of some of > our algorithms, as well as their cache behaviour, on an Itanium2 > platform. However, we don't know how to find out the associated overhead > introduced by using the library. I found one of your presentations > (gelato, May 2005) where you show a sampling overhead by sampling rate > figure. How could we do something similar? > I assume you are talking about sampling measurements. Typically, the bigger the sampling period, the smaller the overhead. But then you need to run for much longer to get a representative set of samples. One way to measure overhead would be to measure a baseline, i.e., the native application without monitoring. Then start with a high sampling period, and slowly make it smaller. At each run measure execution time. You should see that the smaller the period, the longer the run. Which period to start with depends on: the event, the workload. ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ perfmon2-devel mailing list perfmon2-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/perfmon2-devel
