All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
1.0.0 - 2026-03-18
- Transform Kernels: DCT/IDCT for 8x8, 16x16, 32x32 blocks
- Motion Estimation: SAD computation for 4x4, 8x8, 16x16, 32x32, 64x64 blocks
- Loop Filter: Edge-adaptive deblocking filter kernels
- Intra Prediction: DC, Horizontal, Vertical prediction modes
- Device Management: Automatic GPU selection with intelligent scoring
- SYCL Context: Singleton pattern for queue and device management
- CMake Integration: SYCL detection and configuration module
- Unit Tests: Context, transform, and performance tests
- Examples: Basic usage demonstration
- NVIDIA GPUs via CUDA backend
- Intel GPUs via Level Zero backend
- AMD GPUs via HIP backend (experimental)
- CPUs via OpenCL backend (fallback)
- Architecture documentation
- API reference in headers
- Contributing guidelines
- Python bindings
- FFmpeg integration
- OpenCV integration
- Docker images
- vcpkg/conan packages
- Performance benchmarks dashboard
| Version | Date | Highlights |
|---|---|---|
| 1.0.0 | 2026-03-18 | Initial release with core kernels |