- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Is there any way to get the number of cache misses in my program without using vtune?
Maybe a compiler option like -profile-something .. ?
Link Copied
1 Reply
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Well, if you can find something to program the PMU counters (I don't have one - used SEP while it was being distributed), then the _rdpmc intrinsic will let you programmatically query them (recall that only two counters can be active at once in KNC). Note that while we can get decent L1 hit data for KNC, there is no reliable way to get L2 hit data. See http://software.intel.com/en-us/articles/optimization-and-performance-tuning-for-intel-xeon-phi-coprocessors-part-2-understanding for more information.
Cheers,
Charles
![](/skins/images/A813B21DBC25E22B34567C634318FFFF/responsive_peak/images/icon_anonymous_message.png)
Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page