Module simd 

Generic SIMD-aware vector operations.

Public operations are expressed once over sealed scalar traits. The concrete scalar type is selected at the call site, and monomorphization removes the trait layer before execution. Native ISA kernels are private backend details.

Traits

SimdReal

Scalar contract for native-precision real-valued statistics.

SimdScalar

Native-precision scalar contract for SIMD-aware slice operations.

Functions

add

Adds two slices into result.

dot

Computes a native-precision dot product.

has_avx2_support

Returns true when the current x86 CPU supports AVX2.

has_native_vector_path

Returns whether T has a native vector backend on the current CPU.

has_neon_support

Returns true when the current AArch64 CPU supports NEON.

matrix_mul_square

Multiplies two row-major square matrices.

mean

Computes a native-precision arithmetic mean.

mul

Multiplies two slices into result.

sum

Computes a native-precision sum.

variance

Computes a native-precision population variance.