Module simd 

SIMD-accelerated operations for CPU backend.

This module provides high-performance implementations of common GPU-like operations using SIMD (Single Instruction, Multiple Data) instructions.

Operations

• Vector Operations: SAXPY, dot product, element-wise operations
• Reductions: Sum, min, max, mean
• Stencil Operations: 2D/3D Laplacian for FDTD simulations
• Array Operations: Fill, copy, compare

Example

use ringkernel_cpu::simd::SimdOps;

// SAXPY: y = a * x + y
let x = vec![1.0f32; 1024];
let mut y = vec![2.0f32; 1024];
SimdOps::saxpy(2.0, &x, &mut y);

// Reduction
let sum = SimdOps::sum_f32(&y);

Structs

SimdOps

SIMD-accelerated operations.