Struct KernelFactory 

pub struct KernelFactory { /* private fields */ }

Kernel factory for creating optimized kernels

Implementations

impl KernelFactory

pub fn new( arch: String, compute_capabilities: Vec<(i32, i32)>, available_memory: usize, shared_memory_per_block: usize, max_threads_per_block: usize, ) -> Self

Create a new kernel factory

pub fn create_fft_kernel( &self, input_size: usize, input_address: usize, output_address: usize, ) -> FFTResult<FFTKernel>

Create an FFT kernel optimized for the target GPU

pub fn create_sparse_fft_kernel( &self, input_size: usize, sparsity: usize, input_address: usize, output_values_address: usize, output_indices_address: usize, algorithm: SparseFFTAlgorithm, window_function: WindowFunction, ) -> FFTResult<SparseFFTKernel>

Create a sparse FFT kernel optimized for the target GPU

pub fn check_memory_requirements( &self, total_bytesneeded: usize, ) -> FFTResult<()>

Check if there’s enough memory for the requested operation

Trait Implementations

impl Clone for KernelFactory

fn clone(&self) -> KernelFactory

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for KernelFactory

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl KernelFactoryExt for KernelFactory

fn can_use_tensor_cores(&self) -> bool

Check if the GPU can use tensor cores

fn get_optimal_config( &self, signal_size: usize, algorithm: SparseFFTAlgorithm, window_function: WindowFunction, ) -> KernelConfig

Get optimal configuration for a specific algorithm and signal size