- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
We build an application using v11 of your Fortran compiler which makes use of OpenMP. We run the exe with 1 thread and it takes around 6 mins. We re-compile and remove /QopenMP and re-run it and it takes around 6 mins. So, the use of OpenMP has a large overhead in the case when the code is run a single core machine. Is this expected?
Thanks!
Tony
Thanks!
Tony
Link Copied
2 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
You've got "6 mins" twice, which implies no overhead. Is this what you intended?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Oh - sorry about that - here is what I meant to say:
We build an application using v11 of your Fortran compiler which makes use of OpenMP. We run the exe with 1 thread and it takes around9 mins. We re-compile and remove /QopenMP and re-run it and it takes around 6 mins. So, the use of OpenMP has a large overhead in the case when the code is run a single core machine. Is this expected?
We build an application using v11 of your Fortran compiler which makes use of OpenMP. We run the exe with 1 thread and it takes around9 mins. We re-compile and remove /QopenMP and re-run it and it takes around 6 mins. So, the use of OpenMP has a large overhead in the case when the code is run a single core machine. Is this expected?
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