Hi all,
it's far past my bed time, so I'm going to put up a link to some sample code that shows a 5.5x speedup for multiplying long vectors of complex numbers using NEON as opposed to the default libc implementation. Here it is. This has been an "educational" experience in familiarizing myself with the NEON instruction set. Tomorrow, I'm moving on to more FFTW related things, and will finally commit some demo code, and FFTW changes to my git repository.
No comments:
Post a Comment