- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi all
I am doing some experiment on auto-parallelization tools on NAS benchmark, by using Intel icc V18 (FLAGS : -parallel -g -O2 ), even by changing the number of threads (export OMP_NUM_THREADS=4, 8 or 16) always I got the same execution time (it happened for all NAS benchmarks and not just for some of them) !!!
should I use any additional flag ???
Thanks
- Tags:
- CC++
- Development Tools
- Intel® C++ Compiler
- Intel® Parallel Studio XE
- Intel® System Studio
- Optimization
- Parallel Computing
- Vectorization
Link Copied
2 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Do you see the different execution time if you compile with GCC for 4, 8, 16 threads?
Thanks,
Viet
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
You should look at the optimization reports to find out if the automatic parallelization was actually able to parallelize the important loops....

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