Struct CudaTileBackend 

pub struct CudaTileBackend { /* private fields */ }

CUDA tile GPU backend.

Manages CUDA device, compiled kernels, and provides the TileGpuBackend implementation.

Implementations

impl CudaTileBackend

pub fn new(tile_size: u32) -> Result<Self>

Create a new CUDA tile backend.

Compiles the FDTD kernels using NVRTC.

pub fn with_device(ordinal: usize, tile_size: u32) -> Result<Self>

Create a CUDA tile backend on a specific device.

pub fn buffer_width(&self) -> u32

Get buffer width (tile_size + 2 for halos).

pub fn buffer_size_bytes(&self) -> usize

Get buffer size in bytes (18x18 floats for 16x16 tile).

pub fn halo_size_bytes(&self) -> usize

Get halo buffer size in bytes (16 floats).

pub fn interior_size_bytes(&self) -> usize

Get interior buffer size in bytes (16x16 floats).

pub fn device(&self) -> &CudaDevice

Get device reference.

Trait Implementations

impl TileGpuBackend for CudaTileBackend

type Buffer = CudaBuffer

GPU buffer type for this backend.

fn create_tile_buffers( &self, tile_size: u32, ) -> Result<TileGpuBuffers<Self::Buffer>>

Create GPU buffers for a tile.

fn upload_initial_state( &self, buffers: &TileGpuBuffers<Self::Buffer>, pressure: &[f32], pressure_prev: &[f32], ) -> Result<()>

Upload initial pressure data (one-time, at tile creation).

fn fdtd_step( &self, buffers: &TileGpuBuffers<Self::Buffer>, params: &FdtdParams, ) -> Result<()>

Execute FDTD step entirely on GPU (no host transfer).

fn extract_halo( &self, buffers: &TileGpuBuffers<Self::Buffer>, edge: Edge, ) -> Result<Vec<f32>>

Extract halo from GPU to host (small transfer for K2K).

fn inject_halo( &self, buffers: &TileGpuBuffers<Self::Buffer>, edge: Edge, data: &[f32], ) -> Result<()>

Inject neighbor halo from K2K message into GPU buffer.

fn swap_buffers(&self, buffers: &mut TileGpuBuffers<Self::Buffer>)

Swap ping-pong buffers (pointer swap, no data movement).

fn read_interior_pressure( &self, buffers: &TileGpuBuffers<Self::Buffer>, ) -> Result<Vec<f32>>

Read full tile pressure for visualization.

fn synchronize(&self) -> Result<()>

Synchronize GPU operations.