Up to 200x Faster Inner Products and Vector Similarity — for Python, JavaScript, Rust, and C, supporting f64, f32, f16 real & complex, i8, and binary vectors using SIMD for both x86 AVX2 & AVX-512 and Arm NEON & SVE 📐
APACHE-2.0 License
Published by ashvardanian about 1 year ago
cdist
implementations with OpenMP (293504a)SQRT
macro to RSQRT
(d6b6cee)false
(5a66779)SIMSIMD_SQRT
param to avoid LibC (196d679)i8
vectors with integers (a7735aa)_Pragma
parameterized macros (e33e9f7)_mm_mask_rsqrt14_pd
(61e401c)rsqrt
(ebbdb86)rsqrtss
in AVX2 metrics (47fac21)vrsqrte_f32
on Arm NEON (a5a868c)avx512bw
flag (bcba790)Published by ashvardanian about 1 year ago
avx512bw
label for masked loads (ccce7c8)avx512f
flag for __m512
registers (14ceef0)avx512fp16
compilation flag (f8640e0)fma
flag for AVX2 cosine and dot-product (850a0e8)_mm256_popcnt_epi64
(9055a02)union
instantiations (ebc4b12)Published by ashvardanian over 1 year ago
cos_f16_sve
common metric kind for Arm SVE (7b29466)simsimd_cos_i8x16neon
(4dfe9d5)to_int
method for capsules (c0ca719)simsimd_cos_i8x16neon
for Gravitons (ccad0b4)f16
(709e346)size_t
(74760c7)