Module simd 

SIMD-accelerated math primitives for neural forward passes.

Provides AVX2-accelerated dot product, matrix-vector multiply, and activation functions with automatic runtime feature detection and scalar fallback.

On x86_64 targets with the std feature enabled, these functions use runtime is_x86_feature_detected!("avx2") to select AVX2 when available. On all other targets (or without std), they fall back to scalar implementations.

Re-exports

pub use ops::simd_dot;

pub use ops::simd_exp;

pub use ops::simd_mat_vec;

pub use ops::simd_sigmoid;

pub use ops::simd_silu;

pub use ops::simd_tanh;

Modules

ops

Core SIMD-accelerated operations: dot product, matrix-vector multiply, and activation functions.