tasptz / gpuencode

Python and C++ wrapper for encoding videos with the NVidia Video Codec SDK

GPU Encode Library

The purpose of this library is to provide simple access to the NVIDIA Video Codec SDK. The SDK allows to encode high quality videostreams in realtime. Of special interest is the ability to encode videos with H.265 (HEVC) 4K Lossless and 10 bits per channel.

Remark: It utilizes the asynchronous mode only available on Windows!

Install

• install NVIDIA driver and CUDA
• extract NVIDIA Video Codec SDK files to CUDA installation directory or make them available elsewhere
• pip install -r requirements.txt

For the c++ library run

• run build.bat
• package build/gpuencode-*.zip is created

For the python wrapper

• run python setup.py bdist_wheel
• wheel ./dist/gpuencode-*.whl is created
• install with pip install dist/gpuencode-*.whl

Usage

The library provides a simple C++ and python interface, each illustrated by one example.

• for c++ cpp/example_abgr.cpp and cpp/example_concurent.cpp (see example_abgr.exe and example_concurrent.exe inside the build folder)
• for python python/example.py

Supported video cards

https://developer.nvidia.com/video-encode-decode-gpu-support-matrix