google / highway

Performance-portable, length-agnostic SIMD with runtime dispatch
Apache License 2.0
4.13k stars 315 forks source link

Added CeilInt and FloorInt ops #2265

Closed johnplatts closed 3 months ago

johnplatts commented 3 months ago

Added CeilInt and FloorInt ops as AArch64 NEON has instructions for round towards negative infinity (floor) and round towards positive infinity (ceiling) floating point to integer conversions and to provide a more efficient platform-specific implementation of CeilInt/FloorInt on SSE2 and SSSE3.