Dev/eknag/bf16 cossim matmul

ashvardanian / SimSIMD

Up to 200x Faster Dot Products & Similarity Metrics — for Python, Rust, C, JS, and Swift, supporting f64, f32, f16 real & complex, i8, and bit vectors using SIMD for both AVX2, AVX-512, NEON, SVE, & SVE2 📐

https://ashvardanian.com/posts/simsimd-faster-scipy/

Apache License 2.0

988 stars 59 forks source link

Dev/eknag/bf16 cossim matmul #146

Closed eknag closed 4 months ago

eknag commented 4 months ago

Add benchmark - only bf16 cos implementations that leverages vbfmmlaq_f32 instead of 3 seperate dot products.

ashvardanian commented 4 months ago

Thank you, @eknag!