Wire calculatePartialSums to native SIMD via Panama FFI downcall - #651

Closed

r-devulap wants to merge 5 commits into

mainfrom

use-native-calcpartialsum

Commits on May 4, 2026

Add optimal AVX2/SSE reduction and refactor native SIMD code
r-devulap
committed
a9a414f
Wire calculatePartialSums to native SIMD via Panama FFI downcall
r-devulap
committed
a490cc0
perf: Optimize PQ distance look up table calculation for subvector sizes 4,8 & 16 on AVX-512
r-devulap
committed
b530cb7
Optimize size == 2
r-devulap
committed
97a1ada
Remove calculate_partial_sums_f32_512 dispatcher, expose variants directly
r-devulap
committed
0276bf1