HYPE: HYbrid Parallel Encoder
gstreamer, open-source, hypeMaximizing encoding efficiency with HYPE's hybrid parallel architecture.
Hype
The solution to maximize the encoding performance using all your cores
Break the limits of real-time streaming. HYPE uses multi-core parallelization to slash transcoding times and boost resolution without changing your stack.
High-performance parallel encoding
The evolution of the broadcasting and streaming sectors demands higher video resolutions at higher frame rates, while multicore computer architectures are becoming mainstream.
To take advantage of that situation and overcome current limitations 1,2. The project aims to create a hybrid (HW & SW encoders supported) codec-agnostic video encoder that multiplies encoding speeds using all the available encoding cores. To do it, HYPE parallelizes encoding through multiple GPU/CPU cores. HYPE reduces transcoding time for video on demand (VOD) or increases the image resolution for live streaming. It works out of the box with video streaming software and video editors.
High-performance parallel encoding
The evolution of the broadcasting and streaming sectors demands higher video resolutions at higher frame rates, while multicore computer architectures are becoming mainstream.
To take advantage of that situation and overcome current limitations 1,2. The project aims to create a hybrid (HW & SW encoders supported) codec-agnostic video encoder that multiplies encoding speeds using all the available encoding cores. To do it, HYPE parallelizes encoding through multiple GPU/CPU cores. HYPE reduces transcoding time for video on demand (VOD) or increases the image resolution for live streaming. It works out of the box with video streaming software and video editors.
What has been done
Architecture
This innovative architecture reimagines video encoding by introducing parallel processing tailored to specific scenes. Unlike traditional methods that apply uniform encoding settings across an entire video, this design optimizes efficiency and quality by using dedicated encoders for distinct scenes, potentially running on varied configurations.
Integrating a scene detector ensures precise segmentation, while the aggregator guarantees a seamless merger of these segments. In a digital age dominated by video content, this approach offers a faster, more adaptive, and more content-aware encoding strategy, meeting the ever-evolving demands for high-quality video processing.
What has been done
Architecture
This innovative architecture reimagines video encoding by introducing parallel processing tailored to specific scenes. Unlike traditional methods that apply uniform encoding settings across an entire video, this design optimizes efficiency and quality by using dedicated encoders for distinct scenes, potentially running on varied configurations.
Integrating a scene detector ensures precise segmentation, while the aggregator guarantees a seamless merger of these segments. In a digital age dominated by video content, this approach offers a faster, more adaptive, and more content-aware encoding strategy, meeting the ever-evolving demands for high-quality video processing.
How we do it
Methodology
- 01
Scope
Linux Operating System. The encoder is usable through a GStreamer and FFmpeg command line. Test h264, h265 and AV1.
- 02
Proof to concept
We wanted to test different setups mixing encoding types and HW implementations.
- 03
Performance improvement
To assess the limitations of parallel encoding, we gradually increased the number of SW-based encoders.
- 04
Mixed hardware
Run parallel encoding using two vendor boards and test their performance.
What we did
Our achievements
Hybrid encoding validation
MACHINE
- Intel i17 16GB, intel graphics i620.
ENCODING
- Mp4 1920x1080 1h:48m:16s (155904 buffers.)
RESULTS
- HW encoding is slightly slower than HW + SW encoding (scenarios #1 and #3 versus #2), a minor but noticeable difference that validates our approach.
We think the minor improvements are due to a weak laptop processor and, therefore, a bottleneck for CPU decoding.
55% Performance Gain
MACHINE
- AMD Prototype (Ryzen 7 5700G with Radeon Graphics, 16GB).
MACHINE
- Mp4 1920x1080 12m:14s (17616 buffers).
RESULTS
- We could not test the two GPUs simultaneously due to a bug in the AMD drivers. Using the GPU + n·CPU architecture, the encoding time diminishes following an exponential curve, helping us estimate the expected gains for different configurations.
Thus, for this kind of hardware architecture, HYPE is a perfectly viable product and reduces the encoding time by 55% with just seven encoders using GPU+CPU instead of GPU only.
26% Cumulative Speedup
MACHINE
- Intel Core i5-7600K 3.8Ghz, 16GiB RAM, Intel graphics i620 + GeForce GTX 1060 6G.
MACHINE
- Mp4 12m:14s 1280x720 (18350 buffers)
MACHINE
- When comparing the results comparing the improvement between using just the NVidia (18,2 sec), using Nvidia + Intel (15 sec), and Intel + Nvidia + 1 SW encoder (14.3 sec), we observe a limited improvement of 21% using both GPUs and additional 5% when adding a software encoder.
We think that using two equal GPUs will significantly improve the results. Mix HW will not be used on the final product; this is just a PoC to demonstrate our capabilities.
Bits & Bytes
Explore our blog, one byte at a time. Our team unpack our latest news, industry insights and in-depth articles to connect you with the multimedia world.Blog
Read more about our work