Sorry, you need to enable JavaScript to visit this website.

Scalable Video Technology

Scalable Video Technology

Scalable Video Technology (SVT) is a software-based video coding technology that is highly optimized for Intel® Xeon® Scalable processors and Intel® Xeon® D processors. Using the open source SVT encoder core found on this website, it is possible to spread video encoding processing across multiple Intel® Xeon® processors. With many data centers using Intel® Xeon® processors, a real advantage of processing efficiency can be achieved by using a combination of SVT video encoders and Intel® Xeon® processors. The large core count available on modern processors (e.g. up to 56 cores per dual-socket platform) makes it possible to scale the performance of SVT encoders extremely well as a function of the available computational resources. 

A Software-Based Approach to Media Processing

The transcoding of video data remotely in the cloud is experiencing significant growth that is driven by a variety of cloud video applications. As a result, there is a pressing need for a video coding technology that enables encoders to address the many transcoding requirements of such video applications. 

Scalable Video Technology (SVT) is a software-based video coding technology that is highly optimized for Intel® Xeon® Scalable processors and Intel® Xeon® D processors. 

SVT technology and the visual cloud are created to enable this coming wave of new visual applications.

What is the Visual Cloud?

In general terms, the visual cloud refers to the amalgamation of cloud hardware, software, and networking infrastructure that allows efficient remote processing and delivery of media, graphics, and gaming content, as well as enables demanding applications such as media analytics and immersive media. With the ever increasing amount of visual data being generated from various sources, encoding has become a critical part of most visual cloud applications. Encoding is required to compress the source visual content into the least number of bits in the least amount of time, without significantly affecting the visual quality of the source material.

What is SVT?

Scalable Video Technology (SVT) is a software-based video coding technology that allows encoders on any x86 CPU to achieve the best-possible tradeoffs between performance, latency and visual quality. SVT also facilitates performance level scaling of encoders, given the quality and latency requirements of the target applications. The efficiency and scalability of SVT are supported through mainly architectural and algorithmic features, and also via specific optimizations for Intel® Xeon® Scalable processors and Intel® Xeon® D processors.

SVT encoder core as part of a transcoding solution. The SVT encoder core is shown in white. The rest of the visual transcoder’s components are provided by partner codec ISVs and/or customers.

With most of the data centers using Intel® Xeon® processors, targeted optimizations will increase the processing efficiency of workloads in the visual cloud. The large number of cores available on some of the Intel® Xeon® processors (e.g. up to 56 cores per dual-socket Intel® Xeon® Scalable processor) makes it possible to scale the performance of SVT encoders well as a function of the available computational resources. Moreover, the optimization of the SVT encoders for the Intel® Xeon® memory architecture allows for the efficient execution of memory-demanding components in the SVT algorithms. As such, cloud service providers can leverage their existing infrastructure to deliver optimized workloads. SVT was developed to increase the scalability of the core encoder and improve its tradeoffs between performance and visual quality, particularly for high-resolution video content (e.g., 4K, 8K).
 
The SVT architecture allows for the encoder core to be split into independently operating threads, each thread processing a different segment of the input picture, that run in parallel on different processor cores, without introducing any loss in fidelity. This SVT architecture is standard-agnostic, i.e., it can be applied for the development of encoders that are compliant with different standards. SVT allows any standard-compliant encoder to scale its performance properly in response to the compute and memory constraints, while maintaining a graceful degradation in video quality with increasing performance.
 
The SVT architecture is designed to maximize the performance of an SVT encoder on Intel® Xeon® Scalable processors. It is based on three key ideas: (1) Splitting the encoding operation into a set of independent encoding processes, where partition/mode decisions and normative encoding/decoding are de-coupled, and where the degree of parallelism in each process can be controlled through the use of multi-precision-level features; (2) Adopting a prediction structure that allows for efficient picture-level parallelism; and (3) Splitting each picture into segments and processing multiple segments of a picture in parallel to achieve better utilization of the computational resources with no loss in video quality.
 
The SVT architecture provides 3 dimensional parallelism optimization (process based parallelism, picture based parallelism, and segment based parallelism)  
 

SVT-AV1 encoders are made available to the open source community via a highly-permissive AOM patent and BSD-2 clause license, allowing adopters to reduce the time-to-market and cost-of-ownership of each of their SVT-enabled cloud video transcoding solutions.

SVT and the Open Visual Cloud 

The Open Visual Cloud is an open source project that offers a set of pre-defined reference pipelines for various target visual cloud use cases. These reference pipelines are based on optimized open source ingredients across four core building blocks (encode, decode, inference, and render), which are used to deliver visual cloud services. 
 
Scalable Video Technology is a major component of the Open Visual Cloud and plays a key role in all the Open Visual Cloud Services workflow pipelines. To read more about the Open Visual Cloud and the opportunities for developers, visit the Open Visual Cloud 01.org home page at https://01.org/OpenVisualCloud.com

SVT-HEVC

The SVT-HEVC encoder supports HEVC Main and Main10 profiles (up to Level 6.2) and video input resolutions up to 8Kp60, 4:2:0, 8-bit and 10-bit. The SVT-HEVC encoder includes two modes: Objective-quality (OQ) mode and visual-quality (SQ) mode. The SVT-HEVC encoder was released to the Open Source community in September of 2018. 

SVT-AV1

The SVT-AV1 encoder core supports video input resolutions up to 4Kp60, 4:2:0, 8-bit and 10-bit (HDR), The SVT-AV1 encoder core was released to open source on January 31, 2019.
 

SVT-VP9

The SVT-VP9 Encoder is a VP9-compliant encoder library core. The SVT-VP9 Encoder is being optimized to achieve excellent performance levels currently supporting 10 density-quality presets on a system with a dual Intel® Xeon® Scalable processor targeting real-time encoding of up to two 4Kp60 streams on the Gold 6140 with M8. The SVT-VP9 encoder core was released to open source on February 15, 2019.  

 

Next Steps:

 

 

Project: 
Scalable Video Technology