GV Blog

The Broadcast Data Center: Changing the Way We Think About Broadcast.

The Broadcast Data Center: Changing the Way We Think About Broadcast

Looking at the success stories found in new media companies that are leveraging IT solutions to achieve new levels of agility and capturing increasing market share, you find that the agility they achieved was a direct result of the Modern Data Center technologies they were deploying. They are using distributed, scalable networks and virtualized computation to achieve this level of agility which gives them a competitive edge versus the traditional SDI based broadcasters of today. The Broadcast Data Center is a concept developed to characterize new, innovative solutions evolving in the broadcast market.

By adopting a Data Center model using commercial off the shelf (COTS) IP devices the new media companies, such as Netflix, Hulu and Google, enjoy agility of service deployment, scalability, efficient use of computational resources, simple upgrades and an overall non-blocking architecture with graceful fault tolerance. All of these benefits could be leveraged by the broadcast market.

So why weren't broadcasters actively leveraging COTS IP equipment for live production?

Well, live production poses some serious constraints that were not addressed in a typical data center using COTS IP; specifically, having to contend with high persistent bandwidth associated with uncompressed or lightly compressed video. These workflows would stress a typical network with gigabits per second per signal. On top of that, broadcast workflows demand that these large signals are handled with extremely low latency and deterministic delivery as required for real time video applications. Furthermore, video applications often require intensive processing, such as HEVC up-down-cross conversion that requires greater computational intensity than can typically be achieved without dedicated firmware. Finally, the video specific ability to perform vertically accurate switching of video signals was not something COTS IP switches were aware of or were capable of handling.

So, what happens when we take the benefits of COTS IP and merge it with broadcast requirements?

We get the Broadcast Data Center. Grass Valley has taken the advantages that make IT data centers based on COTS IP so alluring and applied them to a Broadcast-centric workflow that meets broadcast constraints. In looking deeper at a typical data center distributed architecture, we felt we could add some true innovation at the edge of network in order to overcome some of the highlighted broadcast specific issues. We created processing nodes we call GV Node, highlighting its role in the larger switching fabric.

These edge processing modules support IP and SDI I/O, IP routing and processing, and IP multiviewer cards. The frame can also be populated with all the existing Densité baseband modules for full IP and baseband interoperability. This is important as formats change or as the facility needs changes as more devices become IP — blades can simply be swapped out or moved to other areas. They also play an important role providing interconnection into the IP fabric for legacy formats, such as SDI, MADI, AES, even analog audio and video as well to provide the processing capability still required in a broadcast production facility or mobile truck in order to provide a finished product.

Another requirement for the Broadcast Data Center is a mechanism for vertically accurate switching, which is critical for live production as we know it. This ability is needed when doing camera shading and having to switch back and forth quickly between two router sources, or when the outer is the backup for the production switcher in case of a failure. This ability to switch in the vertical is also required when switching to an encoder in the case of a direct playout feed or when a replay system using feeds off of a router needs an uninterrupted signal. GV Node not only performs low latency switching in the vertical, but also cleanly switched audio to avoid loud pops, and can aggregate signals for distribution across nodes or the COTS IP fabric.

Implementing a Broadcast Data Center leverages your COTS investment in terms of both switching and compute power, addresses key broadcast needs such as persistent high bandwidth signal management, low latency transport and vertically accurate switching and ensures interoperability by using open SMPTE standards. At the same time, the Broadcast Data Center delivers the agility needed and desired for competiveness in the IP era and paves the way for a fully virtualized broadcast ecosystem similar to what is found in data centers and new media companies today.

The Broadcast Data Center utilizing COTS IP is here today — AND it is certainly changing the way we think about broadcast.

- Sara

Further Reading: