|Message: Xeon Phi performance||Not Logged In (login)|
Click on the Forum title, e.g. on the "Forums by Category" page, to read a sequence of postings to the Forum and its threads all in one page. If you are only interested in one thread or the thread following a specific posting, click the thread or the posting, which takes you to a smaller page, which contains only the part you are interested in and may be easier to navigate.
Messages are "chained" if there are only replies at the first level, i.e. 1/1.html, 1/1/1.html etc. In case of "chained" messages the message number is replaced by the icon and there is no indentation.
Inline: Display the subject line only or also the text of the posting(s); for the choice "All" the "Outline" choices are switched off.
|1||0||1||no text / full text of posting|
|2||1||All||text for level 1 only / text for All postings|
Outline: Choose the depth of the posting thread, successive toggle controls provide increasing detail.
|1||2||1||2 levels / 1 level (original posting)|
|2||3||2||3 levels / 2 levels|
|3||3||All||3 levels / all levels (all postings)|
I have been running Geant4 on a system with an
Intel® Xeon(R) CPU E5-1650 v2 @ 3.50GHz × 12
and a Xeon Phi coprocessor with 6GB of memory and 57 cores which I believe is a 3120a.
I have compiled and ran a simple application on both the CPU and on the MIC. For 1 billion events I find that the run time is
CPU: 72 minutes
MIC: 137 minutes
I expected the performance of the MIC to be comparable to a dual Xeon setup. I also noticed that for 100 million events the time of execution on the MIC was
57 threads: 18 minutes
228 threads: 17 minutes
I am surprised that the performance does not scale better from 57 to 228 threads. I have used Ganglia to monitor the performance of the MIC and the memory consumption is much less than 6GB.
Do you have any suggestions for how to improve the performance on the Xeon Phi?
|Inline Depth:||Outline Depth:||Add message:|