![]() We can filter the list down to h.264 encoders with “grepĢ64”. If NVENC and libx264 built properly you should get ~/Development/ffmpeg_build $ make -j 4Ĭheck that FFmpeg works. ffmpeg_NVIDIA_gpu_acceleration.patchĬonfigure FFmpeg with NVENC, NVRESIZE and x264 support. ~/Development/ $ wget Īpply the NVIDIA acceleration patch. ~/Development/ $ git clone git:///ffmpeg.gitĭownload the NVIDIA acceleration. Get x264 ~/Development $ git clone git:///x264.gitĬonfigure x264 ~/Development/x264 $. ![]() ~/Development/ $ unzip cudautils.zipīuild the CUDA utility. Install the build infrastructure packages sudo apt-get install build-essential git yasm nasm unzip wget sysstat NVENC SDKĬopy the NVENC SDK that you downloaded from above side to ~/Development mkdir DevelopmentĬp Video_Codec_SDK_8.0.14.zip Development/ (or copy paste though GUI)Ĭopy the NVENC headers to /usr/local/include to make it easier later sudo cp Video_Codec_SDK_8.0.14/Samples/common/inc/*.h /usr/local/include CUDA UTILITYĭownload and install a light-weight library to communicate with the CUDA displayĬopy the CUDA utility to ~/Development/. Get NVENC SDK from the NVIDIA developer’s site for cuda enabled ffmpeg libav įollow the steps provided in following document: FFMPEG-with-NVIDIA-Acceleration-on-Ubuntu_UG_v01.pdf The following code will remove ffmpeg and related packages: sudo apt-get -y remove ffmpeg x264 libx264-dev GPU hardware accelerator engine for video decoding (referred to as NVDEC) supports faster than real-time decoding which makes it suitable to be used for transcoding applications, in addition to video playback applications. With decoding/encoding offloaded, the graphics engine and the CPU are free for other operations. NVIDIA GPUs contain one or more hardware-based decoder and encoder(s) (separate from the CUDA cores) which provides fully-accelerated hardware-based video decoding and encoding for several popular codecs.
0 Comments
Leave a Reply. |