Loqid - Composable Disaggregated Video Infrastructure for Video

Scalable Video Encoding Pools for the Data Center

SOLUTION HIGHLIGHTS

Get up to 96 Broadcast Quality 1080p60 Live Streams in 2RU

Ultra High Density

Twenty times increase in video encoding density as compared to software.

8K/4K/UHDTV

Supports a wide variety of streaming applications

AV1, HEVC, H.264

Multi-format Transcoding, Encoding, and Decoding.

AI Deep Neural Network Engines

Enables advanced processing including object detection, classification, segmentation and ROI for image quality improvement and content adaptive rate control.

Ultra Low-Latency Encoding

Enables Interactive video applications including Cloud Mobile Gaming, AR and VR.

Real-Time Encoding

Optimized for live streaming and interactive video applications.

Scalable

High capacity encoding throughput for rapid deployment of additional channels.

Video 2D Processing Engines

Video cropping, padding and scaling for encoding ladder generation and image composition, video overlay, YUV and RGB conversion.

20x higher channel density than software encoding on CPU

Composable Disaggregated Infrastructure

With Liqid’s Composable Disaggregated Infrastructure (CDI), ASIC-based video encoding is as flexible as software running on general-purpose CPUs. 

Say goodbye to hardware walled gardens where video encoding operations are limited by the availability of instances containing the necessary hardware.

Using Liqid technology, any machine on the network can access NETINT’s purpose-built Video Processing Units, allowing hyper-scale video platforms, social networks, and public clouds to provision video encoding and processing resources to meet their performance and capacity requirements without limitation.

Composable Disaggregated Video Encoding Infrastructure

How it works
Composable Disaggregated Video Encoding Infrastructure

Composable Disaggregated Video Encoding technology opens up hardware-based video solutions to be as flexible and available as software by allowing any machine in the data center to natively access all VPUs installed on the network regardless of their physical location.

The number of hardware encoders that can be physically installed into a chassis is often a limiting factor to optimally scaling servers for video encoding. With Liqid’s Composable, Disaggregated Infrastructure, hardware-based video encoders are not installed in the server chassis but are instead disaggregated and placed into external PCIe enclosures, called expansion chassis that are available as compute resources.

Video platforms can have as many expansion chassis as are required to hold their video encoding resources. Interconnecting all the video encoding resources to be composed is accomplished with a fabric, providing every resource direct access to each other. For connections between chassis and racks to be transparent to video encoding workloads, Liqid leverages high bandwidth technology, including PCIe (Gen 4 and Gen 3) and 100GbE.

Once resources are disaggregated and connected over a distributed fabric, Liqid’s composable software is used to create bare metal servers composed with video encoding resources tuned to meet any workload requirement in seconds.

This unique technology enables large-scale video streaming services and platforms to adopt NETINT VPUs and benefit from a greater than 20x cost and density advantage with ASICs without giving up operational flexibility and agility.

Benefits

A new level of flexibility for hardware-based video solutions

Liqid’s Composable, Disaggregated Infrastructure brings a new level of flexibility for hardware-based video solutions by enabling any machine in the data center to natively access all VPUs installed on the network regardless of their physical location. This flexible architecture greatly expands access to advanced video processing functions found only in NETINT’s hardware-based VPUs, enabling greater flexibility, efficiency, and scalability across any video platform.

The unpredictability of stream sessions and the widely varying number of viewers and participants in applications including cloud gaming, live events, and social live video requires video platform operators to have the ability to dynamically allocated video encoding resources for every job, making Liqid’s CDI an ideal solution for dynamic encoding environments.