simvx.graphics.renderer.particle_compute

GPU compute shader particle simulation.

Dispatches a compute shader to update particle positions, velocities, lifetimes, and visual properties entirely on the GPU — avoiding per-frame CPU-to-GPU uploads for particle data.

Module Contents

Classes

ParticleCompute

GPU-based particle simulation via Vulkan compute shader.

Data

__all__

log

API

simvx.graphics.renderer.particle_compute.__all__

[‘ParticleCompute’]

simvx.graphics.renderer.particle_compute.log

‘getLogger(…)’

class simvx.graphics.renderer.particle_compute.ParticleCompute(engine: Any)[source]

GPU-based particle simulation via Vulkan compute shader.

Creates a compute pipeline that updates particle state (position, velocity, colour, scale, lifetime) in an SSBO. The same SSBO can be bound by the graphics particle pass for zero-copy rendering.

Initialization

setup(max_particles: int = 65536) → None[source]

Create compute pipeline, SSBO, and descriptor set.

Args: max_particles: Maximum number of particles in the simulation buffer.

dispatch(cmd: Any, dt: float, emitter_config: dict) → None[source]

Dispatch the compute shader to simulate one step.

Args: cmd: Active command buffer (must be outside a render pass). dt: Delta time in seconds. emitter_config: Dict with emitter parameters: - emitter_pos: (x, y, z) - gravity: (x, y, z) - damping: float - initial_velocity: (x, y, z) - velocity_spread: float - start_colour: (r, g, b, a) - end_colour: (r, g, b, a) - start_scale: float - end_scale: float - emission_radius: float

get_particle_ssbo() → Any[source]

Return the particle SSBO buffer handle for use by the rendering pass.

render(cmd: Any, particle_pass: Any, view_proj: numpy.ndarray, camera_right: numpy.ndarray, camera_up: numpy.ndarray, extent: tuple[int, int]) → None[source]

Draw the GPU-simulated particles using the shared billboard pipeline.

Reuses the same graphics pipeline as ParticlePass (identical particle.vert / particle.frag shaders) but binds our own compute- owned SSBO in place of the pass’s CPU-uploaded one. A lazy-allocated descriptor set matching ParticlePass._ssbo_layout points at the compute buffer so no per-frame upload is needed.

get_particle_memory() → Any[source]

Return the particle SSBO memory handle.

property max_particles: int[source]
property ready: bool[source]
upload_initial_particles(particles: numpy.ndarray) → None[source]

Seed the GPU buffer with CPU-generated particle data.

Args: particles: Numpy array with dtype matching PARTICLE_DTYPE. Length must not exceed max_particles.

cleanup() → None[source]

Destroy all GPU resources.