- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
HI All,
I'm using spbsvx to solve a large band matrix.
While searching for ways to imrpove my SW performance I noticed that spbsvx is not paralleled at all and utilizes only once core of the CPU (i7), even after setting MKL_Set_Num_Threads(8) and MKL_Set_Dynamic(FALSE) (needless to say I also tried with set_dynamic(TRUE).
After a bit more reading I realized that spvsvx is not in the list of LAPACK threaded functions.
Is there a good alternative you can recommend to solve the matrix faster, using more cores?
10x,
Ariel.
Link Copied
0 Replies
![](/skins/images/406383E7D32C62C6E4F7327A35AB307E/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