topic ok, thanks for the update. We in Intel® Integrated Performance Primitives

IPP complex sparse fir filter

Baccouche__Bessem — Wed, 06 Nov 2019 11:23:52 GMT

hello together, I can not find a complex implementation of the sparse fir filter in the ipp,

why is this implementation missing?

will u offer an implementation in the future?

thank you

Bessem

Do you need to see support

Gennady_F_Intel — Thu, 07 Nov 2019 03:37:19 GMT

Do you need to see support 32fc for the FIRSparse function or 64fc also?

These data types were missing because we didn't see such requests in the past. You may submit the FR to the Intel Online Service Center.

thank you for the quick

Baccouche__Bessem — Fri, 08 Nov 2019 05:27:23 GMT

thank you for the quick response, 32fc satisfies my requirements

ok, thanks for the update. We

Gennady_F_Intel — Mon, 11 Nov 2019 05:45:33 GMT

ok, thanks for the update. We will consider the opportunity of this implementation and will keep this thread updated.

Bessem,

Gennady_F_Intel — Wed, 01 Apr 2020 10:02:38 GMT

Bessem,

This API has been added to IPP 2020.1 which is already available for download.

Could you take this update and let us know how it will work on your side, please.

Here is from ipps.h :

/* /////////////////////////////////////////////////////////////////////////////
// Names: ippsFIRSparse
// Purpose: FIRSparse filter with float or complex float taps. Vector filtering
// Parameters:
// pSrc - pointer to the input vector
// pDst - pointer to the output vector
// len - length data vector
// pState - pointer to the filter state
// Return:
// ippStsNullPtrErr - pointer(s) to the data is NULL
// ippStsSizeErr - length of the vectors <= 0
// ippStsNoErr - otherwise
*/
IPPAPI(IppStatus, ippsFIRSparse_32f, (const Ipp32f* pSrc, Ipp32f* pDst, int len, IppsFIRSparseState_32f* pState))
IPPAPI(IppStatus, ippsFIRSparse_32fc, (const Ipp32fc* pSrc, Ipp32fc* pDst, int len, IppsFIRSparseState_32fc* pState))

Gennady