Showing results for

- Intel Community
- Software
- Software Development SDKs and Libraries
- Intel® oneAPI Math Kernel Library
- performance difference of DGEMM on matrix containing huge values

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-16-2012
12:33 AM

156 Views

performance difference of DGEMM on matrix containing huge values

Dear MKL experts,

I've met an issue on matrix multiplication DGEMM routine that if the matrix cotnains huge values, i.e. 1.0d+17, the performance dropped significantly as low as 1/3 of regular scenario.

Any idea if this is possible and/or what the reason could be? I really don't expect such behavior. BTW, this is encountered only in Windows but not in Linux.

thanks,

Link Copied

20 Replies

Gennady_F_Intel

Moderator

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-16-2012
03:23 AM

152 Views

Hi, thanks for the issue.
Is that 32 or 64 bit systems?

Gennady_F_Intel

Moderator

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-16-2012
03:31 AM

152 Views

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-16-2012
04:21 AM

152 Views

Gennady_F_Intel

Moderator

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-16-2012
05:12 AM

152 Views

Shane_S_Intel

Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-16-2012
08:22 AM

152 Views

It would also be interesting to know if there are any +/-Infs in your output.

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-16-2012
07:20 PM

152 Views

Shane_S_Intel

Employee

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-16-2012
09:29 PM

152 Views

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-16-2012
10:07 PM

152 Views

TimP

Black Belt

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-17-2012
05:59 AM

152 Views

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-17-2012
09:28 PM

152 Views

TimP

Black Belt

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-18-2012
04:54 AM

152 Views

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-18-2012
07:08 AM

152 Views

Gennady_F_Intel

Moderator

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-18-2012
09:29 PM

152 Views

Can you give us the example of this case to check the problem on our side?

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-19-2012
11:29 PM

152 Views

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-21-2012
12:20 AM

152 Views

Gennady_F_Intel

Moderator

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-21-2012
02:20 AM

152 Views

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-21-2012
02:47 AM

152 Views

can you let me know where you put /Qftz? compiler option or linker option. details are appreciated.

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-21-2012
02:54 AM

152 Views

TimP

Black Belt

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-21-2012
04:35 AM

152 Views

boreas

Beginner

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

11-22-2012
01:08 AM

58 Views

Thanks, TimP. This really helps understanding. thank you all.

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

For more complete information about compiler optimizations, see our Optimization Notice.