Expand description
GPU acceleration for sparse matrix operations
This module provides GPU acceleration support for sparse matrix operations across multiple backends including CUDA, OpenCL, Metal, ROCm, and WGPU.
Re-exports§
pub use cuda::CudaMemoryManager;pub use cuda::CudaOptimizationLevel;pub use cuda::CudaSpMatVec;pub use metal::MetalDeviceInfo;pub use metal::MetalMemoryManager;pub use metal::MetalOptimizationLevel;pub use metal::MetalSpMatVec;pub use opencl::OpenCLMemoryManager;pub use opencl::OpenCLOptimizationLevel;pub use opencl::OpenCLPlatformInfo;pub use opencl::OpenCLSpMatVec;
Modules§
- convenience
- Convenient functions for common operations
- cuda
- CUDA backend for sparse matrix GPU operations
- metal
- Metal backend for sparse matrix GPU operations on Apple platforms
- opencl
- OpenCL backend for sparse matrix GPU operations
Structs§
- Backend
Info - GPU backend information
- GpuBuffer
- GPU buffer
- GpuContext
- GPU context for managing GPU resources and operations
- GpuDevice
- GPU device abstraction
- GpuKernel
Handle - GPU kernel handle
- GpuSp
MatVec - Unified GPU sparse matrix operations interface
Enums§
- GpuBackend
- GPU backend type
- GpuError
- Error type for GPU operations
- Optimization
Hint - Cross-platform optimization hints
Traits§
- GpuData
Type - Trait for types that can be used with GPU operations