mirror of
https://github.com/keylase/nvidia-patch.git
synced 2024-12-26 22:43:34 +00:00
This patch removes restriction on maximum number of simultaneous NVENC video encoding sessions imposed by Nvidia to consumer-grade GPUs.
.github/ISSUE_TEMPLATE | ||
FFMPEG-with-NVIDIA-Acceleration-on-Ubuntu_UG_v01.pdf | ||
patch.sh | ||
PULL_REQUEST_TEMPLATE.md | ||
README.md |
nvidia-patch
This patch removes restriction on maximum number of simultaneous NVENC video encoding sessions imposed by Nvidia to consumer-grade GPUs.
Requirements:
- x86_64 system architecture
- ubuntu (< 18.04 for 375.39 nvidia driver or kernel < 4.15). Also known to work on Debian and CentOS, but not tested widely.
- nvenc-compatible gpu (https://developer.nvidia.com/video-encode-decode-gpu-support-matrix#Encoder)
- Nvidia driver. Patch availible for:
Tested on Ubuntu 18.04 LTS (GNU/Linux 4.15.0-23-generic x86_64)
step-by-step :
Download driver
https://download.nvidia.com/XFree86/Linux-x86_64/410.78/NVIDIA-Linux-x86_64-410.78.run
Install driver (410.78)
mkdir /opt/nvidia && cd /opt/nvidia
wget https://download.nvidia.com/XFree86/Linux-x86_64/410.78/NVIDIA-Linux-x86_64-410.78.run
chmod +x ./NVIDIA-Linux-x86_64-410.78.run
./NVIDIA-Linux-x86_64-410.78.run
Check driver
nvidia-smi
Patch libnvidia-encode.so (with backup)
bash ./patch.sh
Silent patch libnvidia-encode.so
bash ./patch.sh -s
Rollback libnvidia-encode.so (restore from backup)
bash ./patch.sh -r
See also
If you experience CreateBitstreamBuffer failed: out of memory (10)
, then you have to lower buffers number used for every encoding session. If you are using ffmpeg
, see option -surfaces
("Number of concurrent surfaces") and try value near -surfaces 8
.