Yolanda-Chen
Results
13
comments of
Yolanda-Chen
@fbarchard Please help take a look. Thanks!
> Usually the reads of a C8 kernel out perform a C4 kernel, so I'm suspecting you're spilling registers? For C8 kernels, no spilling after AVX-256 revec, however we cannot...
> Consider c4s2 which is 4 element dot products with rotate of 4 within cell of 8 bytes. Similar to c8, reading 8 channels of source, but instead of 2x...