SIMD-optimized utilities for simulation workloads. More...

#include <immintrin.h>
#include <cstddef>
#include <span>
#include <algorithm>
#include "shm_array.h"

Include dependency graph for shm_simd_utils.h:

This graph shows which files directly or indirectly include this file:

Classes
class	shm_simd::SimdArray< TableType >
	Helper class for SIMD operations on shm_array. More...

Functions
bool	shm_simd::is_aligned (const void *ptr, size_t alignment) noexcept
	Check if pointer is aligned to boundary.

float	shm_simd::sum_floats (const float *data, size_t count) noexcept
	Vectorized sum of float array using AVX2.

float	shm_simd::dot_product (const float a, const float b, size_t count) noexcept
	Vectorized dot product of two float arrays.

void	shm_simd::scale_floats (float *data, size_t count, float scale) noexcept
	Vectorized array scaling (multiply by scalar)

void	shm_simd::fma_floats (const float a, const float b, const float c, float result, size_t count) noexcept
	Vectorized FMA operation: result = a * b + c.

float	shm_simd::min_float (const float *data, size_t count) noexcept
	Find minimum value in float array.

float	shm_simd::max_float (const float *data, size_t count) noexcept
	Find maximum value in float array.

template<int distance = 1>
void	shm_simd::prefetch_read (const void *ptr) noexcept
	Prefetch data for read.

void	shm_simd::prefetch_write (void *ptr) noexcept
	Prefetch data for write.

void	shm_simd::stream_store_floats (float dest, const float src, size_t count) noexcept
	Stream store (bypass cache) for large arrays.

Detailed Description

SIMD-optimized utilities for simulation workloads.

Provides SIMD-friendly operations for common simulation patterns. Requires C++23 and x86-64 with AVX2 or better.

Definition in file shm_simd_utils.h.

Classes