Community
cancel
Showing results for 
Search instead for 
Did you mean: 
markpark
Beginner
74 Views

Preferred Sort Routine

We are making some code parallel which contains qsort(). On a recent Web seminar here, I heard that the standard template library sort was faster. It definitely was faster. When we made the code parallel,VTune pointed to this revised code as containing a LOCK which keeps it the stl code from actually running in parallel efficiently.

Is there a better sort available (linking Microsoft Thread Safe libraries)? Is it possible this goes away in Release version?

0 Kudos
0 Replies
Reply