topic DivCRev_64f in Intel® Integrated Performance Primitives

DivCRev_64f

Ben_Young — Wed, 22 Sep 2010 08:55:11 GMT

Hi,

I was wondering the best place to ask about omissions/desired functionality in the IPP? We're looking to make use of the IPP in our app, and most of the primitives are covered, but I can't find a DivCRev for double precision vectors. Is there a technical reason this is missing, or does it need someone to ask for it?

Thanks,

Ben Young

DivCRev_64f

Naveen_G_Intel — Wed, 22 Sep 2010 09:03:18 GMT

Hi Ben,

This is right forum to ask any questions related to IPP.

IPP has got function to Divides a constant value by each element of a ippsDivCRev.

Please refer to IPP signal processing manual to get additional information about this function.

Regards,

Naveen Gv

DivCRev_64f

Ben_Young — Wed, 22 Sep 2010 09:05:52 GMT

Hi Naveen,

Thanks for the reply. I have seen ippsDivCRev, but there doesn't appear to be a 64f variant! Is that just a doc omission, or is there really no such method?

Thanks,

Ben

DivCRev_64f

Naveen_G_Intel — Wed, 22 Sep 2010 09:17:57 GMT

Hi,

64f variant is not there, means not implemented in IPP. You can try to use ippsDivCRev_32f.

Thanks,

Naveen Gv

DivCRev_64f

Ben_Young — Wed, 22 Sep 2010 09:21:35 GMT

Unfortunately we can't use the 32f variant. We can work around it fine (invert and then mult by constant), but is there a way to request for it to be included in future versions?

Thanks,

Ben