Struct FieldSystemGpu 

pub struct FieldSystemGpu {
    pub fields: Vec<SingleFieldGpu>,
    pub params_buffer: Buffer,
    pub field_count: usize,
    pub merge_pipeline: ComputePipeline,
    pub merge_bind_group_layout: BindGroupLayout,
    pub blur_decay_pipeline: ComputePipeline,
    pub blur_decay_bind_group_layout: BindGroupLayout,
    pub clear_pipeline: ComputePipeline,
    pub clear_bind_group_layout: BindGroupLayout,
}

GPU state for all fields in a simulation.

Fields

fields: Vec<SingleFieldGpu>

Individual field GPU states

params_buffer: Buffer

Parameters buffer (storage buffer with array of FieldParams)

field_count: usize

Number of fields

merge_pipeline: ComputePipeline

Merge pipeline (atomic writes → float field)

merge_bind_group_layout: BindGroupLayout

blur_decay_pipeline: ComputePipeline

Blur/decay pipeline

blur_decay_bind_group_layout: BindGroupLayout

clear_pipeline: ComputePipeline

Clear pipeline (reset atomic buffers to zero)

clear_bind_group_layout: BindGroupLayout

Implementations

impl FieldSystemGpu

pub fn new(device: &Device, registry: &FieldRegistry) -> Self

pub fn process( &mut self, device: &Device, encoder: &mut CommandEncoder, _queue: &Queue, time: f32, delta_time: f32, )

Run field processing: merge deposits, blur/decay or custom update, clear write buffer

If a field has custom_update code, it replaces the blur/decay step with the custom shader.

pub fn create_particle_bind_group( &self, device: &Device, layout: &BindGroupLayout, ) -> Option<BindGroup>

Create bind group for particle compute shader access