Hardware Acceleration
It is recommended to update your configuration to enable hardware accelerated decoding in ffmpeg. Depending on your system, these parameters may not be compatible. More information on hardware accelerated decoding for ffmpeg can be found here: https://trac.ffmpeg.org/wiki/HWAccelIntro
Raspberry Pi 3/4
Ensure you increase the allocated RAM for your GPU to at least 128 (raspi-config > Performance Options > GPU Memory).
NOTICE: If you are using the addon, you may need to turn off Protection mode for hardware acceleration.
ffmpeg:
hwaccel_args: preset-rpi-64-h264
Intel-based CPUs (<10th Generation) via VAAPI
VAAPI supports automatic profile selection so it will work automatically with both H.264 and H.265 streams. VAAPI is recommended for all generations of Intel-based CPUs if QSV does not work.
ffmpeg:
hwaccel_args: preset-vaapi
NOTICE: With some of the processors, like the J4125, the default driver iHD doesn't seem to work correctly for hardware acceleration. You may need to change the driver to i965 by adding the following environment variable LIBVA_DRIVER_NAME=i965 to your docker-compose file or in the frigate.yml for HA OS users.
Intel-based CPUs (>=10th Generation) via Quicksync
QSV must be set specifically based on the video encoding of the stream.
H.264 streams
ffmpeg:
hwaccel_args: preset-intel-qsv-h264
H.265 streams
ffmpeg:
hwaccel_args: preset-intel-qsv-h265
AMD/ATI GPUs (Radeon HD 2000 and newer GPUs) via libva-mesa-driver
VAAPI supports automatic profile selection so it will work automatically with both H.264 and H.265 streams.
Note: You also need to set LIBVA_DRIVER_NAME=radeonsi as an environment variable on the container.
ffmpeg:
hwaccel_args: preset-vaapi
NVIDIA GPUs
While older GPUs may work, it is recommended to use modern, supported GPUs. NVIDIA provides a matrix of supported GPUs and features. If your card is on the list and supports CUVID/NVDEC, it will most likely work with Frigate for decoding. However, you must also use a driver version that will work with FFmpeg. Older driver versions may be missing symbols and fail to work, and older cards are not supported by newer driver versions. The only way around this is to provide your own FFmpeg that will work with your driver version, but this is unsupported and may not work well if at all.
A more complete list of cards and ther compatible drivers is available in the driver release readme.
If your distribution does not offer NVIDIA driver packages, you can download them here.
Docker Configuration
Additional configuration is needed for the Docker container to be able to access the NVIDIA GPU. The supported method for this is to install the NVIDIA Container Toolkit and specify the GPU to Docker. How you do this depends on how Docker is being run:
Docker Compose
services:
frigate:
...
image: ghcr.io/blakeblackshear/frigate:stable
deploy: # <------------- Add this section
resources:
reservations:
devices:
- driver: nvidia
device_ids: ['0'] # this is only needed when using multiple GPUs
count: 1 # number of GPUs
capabilities: [gpu]
Docker Run CLI
docker run -d \
--name frigate \
...
--gpus=all \
ghcr.io/blakeblackshear/frigate:stable
Setup Decoder
The decoder you need to pass in the hwaccel_args will depend on the input video.
A list of supported codecs (you can use ffmpeg -decoders | grep cuvid in the container to get the ones your card supports)
V..... h263_cuvid Nvidia CUVID H263 decoder (codec h263)
V..... h264_cuvid Nvidia CUVID H264 decoder (codec h264)
V..... hevc_cuvid Nvidia CUVID HEVC decoder (codec hevc)
V..... mjpeg_cuvid Nvidia CUVID MJPEG decoder (codec mjpeg)
V..... mpeg1_cuvid Nvidia CUVID MPEG1VIDEO decoder (codec mpeg1video)
V..... mpeg2_cuvid Nvidia CUVID MPEG2VIDEO decoder (codec mpeg2video)
V..... mpeg4_cuvid Nvidia CUVID MPEG4 decoder (codec mpeg4)
V..... vc1_cuvid Nvidia CUVID VC1 decoder (codec vc1)
V..... vp8_cuvid Nvidia CUVID VP8 decoder (codec vp8)
V..... vp9_cuvid Nvidia CUVID VP9 decoder (codec vp9)
For example, for H264 video, you'll select preset-nvidia-h264.
ffmpeg:
hwaccel_args: preset-nvidia-h264
If everything is working correctly, you should see a significant improvement in performance.
Verify that hardware decoding is working by running nvidia-smi, which should show ffmpeg
processes:
nvidia-smi may not show ffmpeg processes when run inside the container due to docker limitations.
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 455.38 Driver Version: 455.38 CUDA Version: 11.1 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 GeForce GTX 166... Off | 00000000:03:00.0 Off | N/A |
| 38% 41C P2 36W / 125W | 2082MiB / 5942MiB | 5% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| 0 N/A N/A 12737 C ffmpeg 249MiB |
| 0 N/A N/A 12751 C ffmpeg 249MiB |
| 0 N/A N/A 12772 C ffmpeg 249MiB |
| 0 N/A N/A 12775 C ffmpeg 249MiB |
| 0 N/A N/A 12800 C ffmpeg 249MiB |
| 0 N/A N/A 12811 C ffmpeg 417MiB |
| 0 N/A N/A 12827 C ffmpeg 417MiB |
+-----------------------------------------------------------------------------+
If you do not see these processes, check the docker logs for the container and look for decoding errors.
These instructions were originally based on the Jellyfin documentation.