Intel® Integrated Performance Primitives
Deliberate problems developing high-performance vision, signal, security, and storage applications.
Announcements
FPGA community forums and blogs on community.intel.com are migrating to the new Altera Community and are read-only. For urgent support needs during this transition, please visit the FPGA Design Resources page or contact an Altera Authorized Distributor.

SVD

Rob_Ottenhoff
New Contributor I
954 Views
Hi!

From other posts I understand that the speach recognition functions have been removed fromm version 7. Sadly, with them the general pupose function ippsSVD* has also been removed. I have always found it strange that this function was not part of the matrix library, that seems more appropriate than speech recognition. Is there any chance this function will again be part of IPP?

Regards,

Rob
p.s. I use the function to calculate the pseudo-inverse of matrices in time critical code, so the speed of ipp is much wanted.
0 Kudos
2 Replies
Ying_H_Intel
Moderator
954 Views
Hi Rob,

Thanks for your feedback. i recorded your request to IPP track database so IPP engineer team will consider them.

For current workaround, you may refer the below two solutions:
1. http://software.intel.com/en-us/forums/showthread.php?t=86908&o=a&s=lr

2. http://software.intel.com/en-us/forums/showthread.php?t=86101&o=a&s=lr

what is your matrix size? if it is bigger (16x16), then mkl svd is suitable.

Best Regards,
Ying H.
0 Kudos
Ying_H_Intel
Moderator
954 Views

Hi Rob, 

Not sure if it is still interesting, I'm sorry to notify you IPP team decide to no restore the functions, but move all of them to MKL library and some replaced functionality.

and the replacement functions are in MKL , for example, SVD in mkl

Driver Routines for Singular Value Decomposition
Routine Name Operation performed
?gesvd :  Computes the singular value decomposition of a general rectangular matrix.
?gesdd :  Computes the singular value decomposition of a general rectangular matrix using a
divide and conquer method.
?gejsv :  Computes the singular value decomposition of a real matrix using a preconditioned
Jacobi SVD method.
?gesvj  :  Computes the singular value decomposition of a real matrix using Jacobi plane
rotations.

Here are articles : https://software.intel.com/en-us/articles/the-alternatives-for-intel-ipp-legacy-small-matrices-domain  for same code samples.

Thanks

Ying  

 

0 Kudos
Reply