Table of Contents

Public classStatic SimdIntrinsicsVectorOps

Namespace
Rowles.LeanLucene.Search.Simd
Assembly
Rowles.LeanLucene.dll

Explicit-intrinsics SIMD path for vector arithmetic. Selects the widest available instruction set at runtime: AVX-512 → AVX2 → SSE → scalar. Provided alongside SimdVectorOps for benchmark comparison; the slower of the two will be removed once measurements are concluded.

public static class SimdIntrinsicsVectorOps
SimdIntrinsicsVectorOps

Public property IsAvx2Supported

True when the running hardware supports AVX2 (256-bit float path with FMA).

Public property IsAvx512Supported

True when the running hardware supports AVX-512F (single-precision FMA path).

Public property IsSseSupported

True when the running hardware supports SSE (128-bit float path).

Public method CosineSimilarity(ReadOnlySpan<float>, ReadOnlySpan<float>)

Cosine similarity between two equal-length vectors, using explicit intrinsics.

Public method DotProduct(ReadOnlySpan<float>, ReadOnlySpan<float>)

Dot product using explicit intrinsics.

Edit this page