Expand description
GPU Acceleration - CUDA/OpenCL Support
This module provides GPU acceleration utilities for tensor operations. Note: This is a foundational implementation. Full GPU support requires external libraries like wgpu, vulkan, or CUDA bindings.
Structs§
- Auto
Mixed Precision - Automatic Mixed Precision (AMP) support
- Device
Info - GPU device information
- GPUContext
- GPU context manager
- GPUMemory
Manager - Memory management for GPU
- GPUOps
- GPU-accelerated operations
- GPUTensor
- GPU tensor wrapper
Enums§
- Device
Type - GPU device type