파일
rocm-systems/projects/rocdecode/samples/videoDecodePerf/README.md
T
Lakshmi Kumar a1d39cf9f9 Add a sample for measuring the video decoding performance using multi-threads (#50)
* introducing perf example

* chnages for google standards

* adding exception handling

* review comments

* change logic to find vcn devices

* adding comments

[ROCm/rocdecode commit: 5bac02fab3]
2023-11-10 11:23:44 -05:00

37 라인
943 B
Markdown

# Video Decode Sample
This sample illustrates the FFMPEG demuxer to get the individual frames which are then decoded on AMD hardware using VAAPI.
This sample supports both YUV420 8-bit and 10-bit streams.
This sample uses multiple threads to decode the same input video parallely.
## Prerequisites:
* Linux distribution
+ Ubuntu - `20.04` / `22.04`
* [ROCm supported hardware](https://rocm.docs.amd.com/en/latest/release/gpu_os_support.html)
* Install [ROCm 5.5 or later](https://rocmdocs.amd.com/en/latest/deploy/linux/installer/install.html) with `--usecase=graphics,rocm --no-32`
* rocDecode
* CMake `3.5` or later
* [FFMPEG](https://ffmpeg.org/about.html)
```
sudo apt install ffmpeg libavcodec-dev libavformat-dev libswscale-dev
```
## Build
```
mkdir build
cd build
cmake ../
make -j
```
# Run
```
./videodecodeperf -i <input video file [required]>
-t <number of threads [optional - default:4]>
```