Enabling GPU video decoder/encoder¶
TorchAudio can make use of hardware-based video decoding and encoding supported by underlying FFmpeg libraries that are linked at runtime.
Using NVIDIA’s GPU decoder and encoder, it is also possible to pass around CUDA Tensor directly, that is decode video into CUDA tensor or encode video from CUDA tensor, without moving data from/to CPU.
This improves the video throughput significantly. However, please note that not all the video formats are supported by hardware acceleration.
This page goes through how to build FFmpeg with hardware acceleration. For the detail on the performance of GPU decoder and encoder please see Hardware-Accelerated Video Decoding and Encoding
Overview¶
Using them in TorchAduio requires additional FFmpeg configuration.
In the following, we look into how to enable GPU video decoding with NVIDIA’s Video codec SDK. To use NVENC/NVDEC with TorchAudio, the following items are required.
NVIDIA GPU with hardware video decoder/encoder.
FFmpeg libraries compiled with NVDEC/NVENC support. †
PyTorch / TorchAudio with CUDA support.
TorchAudio’s official binary distributions are compiled to work with FFmpeg 4 libraries, and they contain the logic required for hardware-based decoding/encoding.
In the following, we build FFmpeg 4 libraries with NVDEC/NVENC support. If you would like to use FFmpeg 5, then you need to build TorchAudio with it.
The following procedure was tested on Ubuntu.
† For details on NVDEC/NVENC and FFmpeg, please refer to the following articles.
Check the GPU and CUDA version¶
First, check the available GPU. Here, we have Tesla T4 with CUDA Toolkit 11.2 installed.
$ nvidia-smi
Fri Oct 7 13:01:26 2022
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 460.32.03 Driver Version: 460.32.03 CUDA Version: 11.2 |
|-------------------------------+----------------------+----------------------+
| 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 Tesla T4 Off | 00000000:00:04.0 Off | 0 |
| N/A 56C P8 10W / 70W | 0MiB / 15109MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+
Checking the compute capability¶
Later, we need the version of compute capability supported by this GPU. The following page lists the GPUs and corresponding compute capabilities. The compute capability of T4 is 7.5.
Install NVIDIA Video Codec Headers¶
To build FFmpeg with NVDEC/NVENC, we first need to install the headers that FFmpeg uses to interact with Video Codec SDK.
Since we have CUDA 11 working in the system, we use one of n11 tag.
git clone https://git.videolan.org/git/ffmpeg/nv-codec-headers.git
cd nv-codec-headers
git checkout n11.0.10.1
sudo make install
The location of installation can be changed with make PREFIX=<DESIRED_DIRECTORY> install.
Cloning into 'nv-codec-headers'...
remote: Enumerating objects: 819, done.
remote: Counting objects: 100% (819/819), done.
remote: Compressing objects: 100% (697/697), done.
remote: Total 819 (delta 439), reused 0 (delta 0)
Receiving objects: 100% (819/819), 156.42 KiB | 410.00 KiB/s, done.
Resolving deltas: 100% (439/439), done.
Note: checking out 'n11.0.10.1'.
You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by performing another checkout.
If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -b with the checkout command again. Example:
git checkout -b <new-branch-name>
HEAD is now at 315ad74 add cuMemcpy
sed 's#@@PREFIX@@#/usr/local#' ffnvcodec.pc.in > ffnvcodec.pc
install -m 0755 -d '/usr/local/include/ffnvcodec'
install -m 0644 include/ffnvcodec/*.h '/usr/local/include/ffnvcodec'
install -m 0755 -d '/usr/local/lib/pkgconfig'
install -m 0644 ffnvcodec.pc '/usr/local/lib/pkgconfig'
Install FFmpeg dependencies¶
Next, we install tools and libraries required during the FFmpeg build. The minimum requirement is Yasm. Here we additionally install H264 video codec and HTTPS protocol, which we use later for verifying the installation.
sudo apt -qq update
sudo apt -qq install -y yasm libx264-dev libgnutls28-dev
... Omitted for brevity ...
STRIP install-libavutil-shared
Setting up libx264-dev:amd64 (2:0.152.2854+gite9a5903-2) ...
Setting up yasm (1.3.0-2build1) ...
Setting up libunbound2:amd64 (1.6.7-1ubuntu2.5) ...
Setting up libp11-kit-dev:amd64 (0.23.9-2ubuntu0.1) ...
Setting up libtasn1-6-dev:amd64 (4.13-2) ...
Setting up libtasn1-doc (4.13-2) ...
Setting up libgnutlsxx28:amd64 (3.5.18-1ubuntu1.6) ...
Setting up libgnutls-dane0:amd64 (3.5.18-1ubuntu1.6) ...
Setting up libgnutls-openssl27:amd64 (3.5.18-1ubuntu1.6) ...
Setting up libgmpxx4ldbl:amd64 (2:6.1.2+dfsg-2) ...
Setting up libidn2-dev:amd64 (2.0.4-1.1ubuntu0.2) ...
Setting up libidn2-0-dev (2.0.4-1.1ubuntu0.2) ...
Setting up libgmp-dev:amd64 (2:6.1.2+dfsg-2) ...
Setting up nettle-dev:amd64 (3.4.1-0ubuntu0.18.04.1) ...
Setting up libgnutls28-dev:amd64 (3.5.18-1ubuntu1.6) ...
Processing triggers for man-db (2.8.3-2ubuntu0.1) ...
Processing triggers for libc-bin (2.27-3ubuntu1.6) ...
Build FFmpeg with NVDEC/NVENC support¶
Next we download the source code of FFmpeg 4. We use 4.4.2 here. Any version later than 4.1 should work with TorchAudio binary distributions. If you want to use FFmpeg 5, then you need to build TorchAudio after building FFmpeg.
wget -q https://github.com/FFmpeg/FFmpeg/archive/refs/tags/n4.4.2.tar.gz
tar -xf n4.4.2.tar.gz
cd FFmpeg-n4.4.2
Next we configure FFmpeg build. Note the following:
We provide flags like
-I/usr/local/cuda/include,-L/usr/local/cuda/lib64to let the build process know where the CUDA libraries are found.We provide flags like
--enable-nvdecand--enable-nvencto enable NVDEC/NVENC.We also provide NVCC flags with compute capability
75, which corresponds to7.5of T4. †We install the library in
/usr/lib/.
Note
† The configuration script verifies NVCC by compiling a sample code. By default it uses old compute capability such as 30, which is no longer supported by CUDA 11. So it is required to set a correct compute capability.
prefix=/usr/
ccap=75
./configure \
--prefix="${prefix}" \
--extra-cflags='-I/usr/local/cuda/include' \
--extra-ldflags='-L/usr/local/cuda/lib64' \
--nvccflags="-gencode arch=compute_${ccap},code=sm_${ccap} -O2" \
--disable-doc \
--enable-decoder=aac \
--enable-decoder=h264 \
--enable-decoder=h264_cuvid \
--enable-decoder=rawvideo \
--enable-indev=lavfi \
--enable-encoder=libx264 \
--enable-encoder=h264_nvenc \
--enable-demuxer=mov \
--enable-muxer=mp4 \
--enable-filter=scale \
--enable-filter=testsrc2 \
--enable-protocol=file \
--enable-protocol=https \
--enable-gnutls \
--enable-shared \
--enable-gpl \
--enable-nonfree \
--enable-cuda-nvcc \
--enable-libx264 \
--enable-nvenc \
--enable-cuvid \
--enable-nvdec
install prefix /usr/
source path .
C compiler gcc
C library glibc
ARCH x86 (generic)
big-endian no
runtime cpu detection yes
standalone assembly yes
x86 assembler yasm
MMX enabled yes
MMXEXT enabled yes
3DNow! enabled yes
3DNow! extended enabled yes
SSE enabled yes
SSSE3 enabled yes
AESNI enabled yes
AVX enabled yes
AVX2 enabled yes
AVX-512 enabled yes
XOP enabled yes
FMA3 enabled yes
FMA4 enabled yes
i686 features enabled yes
CMOV is fast yes
EBX available yes
EBP available yes
debug symbols yes
strip symbols yes
optimize for size no
optimizations yes
static no
shared yes
postprocessing support no
network support yes
threading support pthreads
safe bitstream reader yes
texi2html enabled no
perl enabled yes
pod2man enabled yes
makeinfo enabled no
makeinfo supports HTML no
External libraries:
alsa libx264 lzma
bzlib libxcb zlib
gnutls libxcb_shape
iconv libxcb_xfixes
External libraries providing hardware acceleration:
cuda cuvid nvenc
cuda_llvm ffnvcodec v4l2_m2m
cuda_nvcc nvdec
Libraries:
avcodec avformat swscale
avdevice avutil
avfilter swresample
Programs:
ffmpeg ffprobe
Enabled decoders:
aac hevc rawvideo
av1 mjpeg vc1
h263 mpeg1video vp8
h264 mpeg2video vp9
h264_cuvid mpeg4
Enabled encoders:
h264_nvenc libx264
Enabled hwaccels:
av1_nvdec mpeg1_nvdec vp8_nvdec
h264_nvdec mpeg2_nvdec vp9_nvdec
hevc_nvdec mpeg4_nvdec wmv3_nvdec
mjpeg_nvdec vc1_nvdec
Enabled parsers:
h263 mpeg4video vp9
Enabled demuxers:
mov
Enabled muxers:
mov mp4
Enabled protocols:
file tcp
https tls
Enabled filters:
aformat hflip transpose
anull null trim
atrim scale vflip
format testsrc2
Enabled bsfs:
aac_adtstoasc null vp9_superframe_split
h264_mp4toannexb vp9_superframe
Enabled indevs:
lavfi
Enabled outdevs:
License: nonfree and unredistributable
Now we build and install
make clean
make -j
sudo make install
... Omitted for brevity ...
INSTALL libavdevice/libavdevice.so
INSTALL libavfilter/libavfilter.so
INSTALL libavformat/libavformat.so
INSTALL libavcodec/libavcodec.so
INSTALL libswresample/libswresample.so
INSTALL libswscale/libswscale.so
INSTALL libavutil/libavutil.so
INSTALL install-progs-yes
INSTALL ffmpeg
INSTALL ffprobe
Checking the intallation¶
To verify that the FFmpeg we built have CUDA support, we can check the list of available decoders and encoders.
ffprobe -hide_banner -decoders | grep h264
VFS..D h264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10
V..... h264_cuvid Nvidia CUVID H264 decoder (codec h264)
ffmpeg -hide_banner -encoders | grep 264
V..... libx264 libx264 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 (codec h264)
V....D h264_nvenc NVIDIA NVENC H.264 encoder (codec h264)
The following command fetches video from remote server, decode with NVDEC (cuvid) and re-encode with NVENC. If this command does not work, then there is an issue with FFmpeg installation, and TorchAudio would not be able to use them either.
$ src="https://download.pytorch.org/torchaudio/tutorial-assets/stream-api/NASAs_Most_Scientifically_Complex_Space_Observatory_Requires_Precision-MP4_small.mp4"
$ ffmpeg -hide_banner -y -vsync 0 \
-hwaccel cuvid \
-hwaccel_output_format cuda \
-c:v h264_cuvid \
-resize 360x240 \
-i "${src}" \
-c:a copy \
-c:v h264_nvenc \
-b:v 5M test.mp4
Note that there is Stream #0:0 -> #0:0 (h264 (h264_cuvid) -> h264 (h264_nvenc)), which means that video is decoded with h264_cuvid decoder and h264_nvenc encoder.
Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'https://download.pytorch.org/torchaudio/tutorial-assets/stream-api/NASAs_Most_Scientifically_Complex_Space_Observatory_Requires_Precision-MP4_small.mp4':
Metadata:
major_brand : mp42
minor_version : 512
compatible_brands: mp42iso2avc1mp41
encoder : Lavf58.76.100
Duration: 00:03:26.04, start: 0.000000, bitrate: 1294 kb/s
Stream #0:0(eng): Video: h264 (High) (avc1 / 0x31637661), yuv420p(tv, bt709), 960x540 [SAR 1:1 DAR 16:9], 1156 kb/s, 29.97 fps, 29.97 tbr, 30k tbn, 59.94 tbc (default)
Metadata:
handler_name : ?Mainconcept Video Media Handler
vendor_id : [0][0][0][0]
Stream #0:1(eng): Audio: aac (LC) (mp4a / 0x6134706D), 48000 Hz, stereo, fltp, 128 kb/s (default)
Metadata:
handler_name : #Mainconcept MP4 Sound Media Handler
vendor_id : [0][0][0][0]
Stream mapping:
Stream #0:0 -> #0:0 (h264 (h264_cuvid) -> h264 (h264_nvenc))
Stream #0:1 -> #0:1 (copy)
Press [q] to stop, [?] for help
Output #0, mp4, to 'test.mp4':
Metadata:
major_brand : mp42
minor_version : 512
compatible_brands: mp42iso2avc1mp41
encoder : Lavf58.76.100
Stream #0:0(eng): Video: h264 (Main) (avc1 / 0x31637661), cuda(tv, bt709, progressive), 360x240 [SAR 1:1 DAR 3:2], q=2-31, 5000 kb/s, 29.97 fps, 30k tbn (default)
Metadata:
handler_name : ?Mainconcept Video Media Handler
vendor_id : [0][0][0][0]
encoder : Lavc58.134.100 h264_nvenc
Side data:
cpb: bitrate max/min/avg: 0/0/5000000 buffer size: 10000000 vbv_delay: N/A
Stream #0:1(eng): Audio: aac (LC) (mp4a / 0x6134706D), 48000 Hz, stereo, fltp, 128 kb/s (default)
Metadata:
handler_name : #Mainconcept MP4 Sound Media Handler
vendor_id : [0][0][0][0]
frame= 6175 fps=1712 q=11.0 Lsize= 37935kB time=00:03:26.01 bitrate=1508.5kbits/s speed=57.1x
video:34502kB audio:3234kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.526932%
Using the GPU decoder/encoder from TorchAudio¶
Checking the installation¶
Once the FFmpeg is properly working with hardware acceleration, we need to check if TorchAudio can pick it up correctly.
There are utility functions to query the capability of FFmpeg in torchaudio.utils.ffmpeg_utils.
You can first use get_video_decoders() and get_video_encoders() to check if GPU decoders and encoders (such as h264_cuvid and h264_nvenc) are listed.
It is often the case where there are multiple FFmpeg installations in the system, and TorchAudio is loading one different than expected. In such cases, use of ffmpeg to check the installation does not help. You can use functions like get_build_config() and get_versions() to get information about FFmpeg libraries TorchAudio loaded.
from torchaudio.utils import ffmpeg_utils
print("Library versions:")
print(ffmpeg_utils.get_versions())
print("\nBuild config:")
print(ffmpeg_utils.get_build_config())
print("\nDecoders:")
print([k for k in ffmpeg_utils.get_video_decoders().keys() if "cuvid" in k])
print("\nEncoders:")
print([k for k in ffmpeg_utils.get_video_encoders().keys() if "nvenc" in k])
Library versions:
{'libavutil': (56, 31, 100), 'libavcodec': (58, 54, 100), 'libavformat': (58, 29, 100), 'libavfilter': (7, 57, 100), 'libavdevice': (58, 8, 100)}
Build config:
--prefix=/usr --extra-version=0ubuntu0.1 --toolchain=hardened --libdir=/usr/lib/x86_64-linux-gnu --incdir=/usr/include/x86_64-linux-gnu --arch=amd64 --enable-gpl --disable-stripping --enable-avresample --disable-filter=resample --enable-avisynth --enable-gnutls --enable-ladspa --enable-libaom --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libcdio --enable-libcodec2 --enable-libflite --enable-libfontconfig --enable-libfreetype --enable-libfribidi --enable-libgme --enable-libgsm --enable-libjack --enable-libmp3lame --enable-libmysofa --enable-libopenjpeg --enable-libopenmpt --enable-libopus --enable-libpulse --enable-librsvg --enable-librubberband --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libssh --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx265 --enable-libxml2 --enable-libxvid --enable-libzmq --enable-libzvbi --enable-lv2 --enable-omx --enable-openal --enable-opencl --enable-opengl --enable-sdl2 --enable-libdc1394 --enable-libdrm --enable-libiec61883 --enable-nvenc --enable-chromaprint --enable-frei0r --enable-libx264 --enable-shared
Decoders:
['h264_cuvid', 'hevc_cuvid', 'mjpeg_cuvid', 'mpeg1_cuvid', 'mpeg2_cuvid', 'mpeg4_cuvid', 'vc1_cuvid', 'vp8_cuvid', 'vp9_cuvid']
Encoders:
['h264_nvenc', 'nvenc', 'nvenc_h264', 'nvenc_hevc', 'hevc_nvenc']
Using the hardware decoder¶
Once the installation and the runtime linking work fine, then you can test the GPU decoding with the following.
For the detail on the performance of GPU decoder and encoder please see Hardware-Accelerated Video Decoding and Encoding
from torchaudio.io import StreamReader
src = "https://download.pytorch.org/torchaudio/tutorial-assets/stream-api/NASAs_Most_Scientifically_Complex_Space_Observatory_Requires_Precision-MP4_small.mp4"
s = StreamReader(src)
s.add_video_stream(
5,
decoder="h264_cuvid",
hw_accel="cuda:0",
decoder_option={
"resize": "360x240",
},
)
s.fill_buffer()
chunk, = s.pop_chunks()
print(' - Chunk:', chunk.shape, chunk.device, chunk.dtype)
- Chunk: torch.Size([5, 3, 240, 360]) cuda:0 torch.uint8
