Expand description
GPU backend implementations and detection utilities
This module contains backend-specific implementations for various GPU platforms and utilities for detecting available GPU backends.
Re-exports§
pub use cuda::get_optimizer_kernels;pub use cuda::CudaContext;pub use cuda::CudaStream;pub use opencl::OpenCLContext;pub use wgpu::WebGPUContext;
Modules§
- cuda
- CUDA backend implementation for GPU operations
- opencl
- OpenCL backend implementation for GPU operations
- wgpu
- WebGPU backend implementation for GPU operations
Structs§
- GpuDetection
Result - Detection results for all available GPU backends
- GpuInfo
- Information about available GPU hardware
Functions§
- check_
backend_ installation - Check if a specific backend is properly installed and functional
- detect_
gpu_ backends - Detect available GPU backends and devices
- get_
device_ info - Get detailed information about a specific GPU device
- initialize_
optimal_ backend - Initialize the optimal GPU backend for the current system