Module gpu_ops 

GPU-accelerated operations for sparse matrices

This module provides GPU acceleration for sparse matrix operations using the scirs2-core GPU backend system. The implementation has been modularized for better maintainability and vendor-specific optimizations.

Re-exports

pub use crate::gpu::BackendInfo;

pub use crate::gpu::GpuSpMatVec;

pub use crate::gpu::OptimizationHint;

pub use crate::gpu::convenience::available_backends;

pub use crate::gpu::convenience::gpu_spmv;

pub use crate::gpu::convenience::gpu_spmv_optimized;

pub use crate::gpu::*;

Structs

AdvancedGpuOps

GpuBuffer

GPU buffer

GpuContext

GPU context for managing GPU resources and operations

GpuKernelHandle

GPU kernel handle

GpuKernelScheduler

GpuMemoryManager

GpuOptions

GpuProfiler

OptimizedGpuOps

SpMVKernel

Enums

GpuBackend

GPU backend type

GpuError

Error type for GPU operations

Traits

GpuBufferExt

GpuDataType

Trait for types that can be used with GPU operations

Functions

gpu_advanced_spmv

Advanced GPU sparse matrix-vector multiplication with optimization hints

gpu_sparse_matvec

GPU sparse matrix-vector multiplication (legacy interface)

gpu_sym_sparse_matvec

GPU symmetric sparse matrix-vector multiplication (legacy interface)