Module simd 

SIMD-accelerated operations using Trueno.

This module provides hardware-accelerated vector and matrix operations for the Presentar rendering pipeline.

When the simd feature is disabled, operations fall back to scalar implementations.

Example

use presentar_core::simd::{Vec4, Mat4, batch_transform_points};
use presentar_core::Point;

let transform = Mat4::identity();
let points = vec![Point::new(0.0, 0.0), Point::new(100.0, 100.0)];
let transformed = batch_transform_points(&points, &transform);

Structs

Mat4

4x4 matrix for transforms.

Vec4

4-component vector for SIMD operations.

Functions

batch_add_simd

SIMD-accelerated batch add using trueno.

batch_dot_product

SIMD-accelerated batch dot product.

batch_lerp_points

Batch linear interpolation.

batch_mean_f64

SIMD-friendly mean of f64 values.

batch_min_max_f64

SIMD-friendly min/max of f64 values.

batch_scale_f64

SIMD-friendly scale operation.

batch_scale_offset_f64

SIMD-friendly scale and offset operation.

batch_stddev_f64

SIMD-friendly standard deviation.

batch_sum_f64

SIMD-friendly sum of f64 values.

batch_transform_points

Batch transform multiple points.

batch_transform_vec4

Batch transform multiple Vec4s.

batch_variance_f64

SIMD-friendly variance calculation (population variance).

best_backend

Get the best available SIMD backend.

bounding_box

Axis-aligned bounding box from points.

centroid

Calculate the centroid of points.

dot_simd

SIMD-accelerated dot product using trueno.

histogram_f64

Compute histogram bin counts for f64 data.

normalize_f64

SIMD-friendly normalization to [0, 1] range.

normalize_with_range_f64

SIMD-friendly normalization with provided range.

percentile_sorted_f64

SIMD-friendly percentile calculation.

point_in_convex_polygon

Check if a point is inside a convex polygon.

polygon_area

Compute the area of a polygon using the shoelace formula.

scale_simd

SIMD-accelerated scale using trueno.

simd_to_vec4

Create Vec4 from SIMD vector.

vec4_to_simd

Create a SIMD-backed vector from Vec4.

weighted_sum_f64

SIMD-friendly weighted sum.

Type Aliases

SimdVectorF32

SIMD vector type from trueno (f32).