CUTLASS 2.10 updates (#622)

Co-authored-by: Aniket Shivam <ashivam@nvidia.com>
This commit is contained in:
ANIKET SHIVAM
2022-09-12 18:26:30 -07:00
committed by GitHub
parent beae168f90
commit e773429f7e
96 changed files with 8365 additions and 1667 deletions

View File

@@ -561,6 +561,17 @@ CUTLASS_HOST_DEVICE int64_t OffsetBytes(int64_t index) {
}
}
CUTLASS_HOST_DEVICE int64_t OffsetBytes(int64_t index, int64_t element_sizeof_bits) {
if (element_sizeof_bits >= 8) {
return index * (element_sizeof_bits / 8);
}
else {
int64_t const kElementsPerByte = ((8 / element_sizeof_bits) + ((element_sizeof_bits >= 8) ? 1 : 0));
return index / kElementsPerByte;
}
}
/////////////////////////////////////////////////////////////////////////////////////////////////
// Min/Max
/////////////////////////////////////////////////////////////////////////////////////////////////