3 Commits

Author	SHA1	Message	Date
Feoramund	c8a62ee4ec	Make simd_util index procs contextless where applicable	2024-08-09 18:54:04 -04:00
Feoramund	12dd0cb72a	Simplify and make simd_util cross-platform ... This new algorithm uses a Scalar->Vector->Scalar iteration loop which requires no masking off of any incomplete data chunks. Also, the width was reduced to 32 bytes instead of 64, as I found this to be about as fast as the previous 64-byte x86 version.	2024-08-09 18:54:04 -04:00
Feoramund	8deeb40e5d	Add vectorized index_byte and last_index_byte	2024-08-06 15:19:01 -04:00