Broadcast Management and Moving to the Cloud

READ MORE: Where Do We Go From Here? (Dell/TV Tech)

The move to cloud is often presented as a binary choice: take advantage of software defined, virtualized infrastructure — perhaps in a public cloud hyperscaler — to drive down operating costs and scale with greater efficiency, or retain existing siloed technology on-premises until every last piece of gear has been amortized and the move to cloud is all or nothing.

According to TV Tech in conjunction with sponsor Dell Technologies, the workflow should determine the technology choice rather than financial decisions dictating the workflow.

In a joint white paper, “Where Do We Go From Here?” they reveal six myths and misconceptions surrounding public cloud hyperscalers in media applications.

It argues that while the resource and operational efficiencies of cloud, IP and IT transport can deliver huge benefits, “there are equally good reasons to tiptoe forward, wringing as much use and value out of in-place technology as possible while making a gradual transition to on-prem or public cloud-based IT alternatives.”

At this point, there is no reason for broadcasters to choose between an on-prem SDI or SDI-IT hybrid architecture and a dedicated public cloud hyperscaler, they say.

“Broadcasters and other M&E enterprises do not have to go all-in on any one public cloud provider — or leverage the public cloud at all — to achieve their operational goals and vision for workflows.”

Those myths and misconceptions about public cloud in turn:

1. The Ultimate Remote Workflow Enabler

It’s agreed that the pandemic accelerated adoption of remote media workflows. The next logical step for many M&E companies appears to be a move from brick-and-mortar production centers to virtualized equivalents in the public cloud.

“However, it’s clear that it’s not the public cloud that is making the evolution in REMI possible but rather the ubiquitous availability of bandwidth to transport live camera feeds and large media files from locations such as sports venues, contributions from reporters in the field and user generated content.”

The same virtualized production tasks that are sometimes deployed in a single public cloud hyperscaler “could just as easily be performed” in an on-prem datacenter with traditional production technology or in a truly multicloud environment. This would enable organizations to extend the life of their existing technology and allowing tactical use of multiple public cloud providers.

2. The Only Path to OPEX

Per the white paper: Broadcast management may seek to use the public cloud to extract their organizations’ workflows “from the seemingly endless cycle of expensive capital outlays, ongoing spending for maintenance, the cost of service contracts, end-of-life depreciation and technological obsolescence.”

Breaking free of the CAPEX approach, it is commonly thought, will save money and reduce all of the expenses associated with maintaining that investment.

However, Dell points to another path. This is the offer from some tech providers (looking at you Dell) of “flexible compute options that enable organizations to realize the benefits of an OPEX-centric business model while avoiding issues that arise with storing valuable media assets in the public cloud.”

3. Exclusive Path to AI/ML

This is a big one. It’s the claim that if media companies want to access all the firepower and smarts of AI tools for anything from file processing to and speech to text data analytics then the best if not only way is via the cloud. Amazon and Microsoft for example both offer AI/ML functions that are pitched as transforming media workflows and making content more valuable.

“The danger lies in painting the public cloud-AI/ML canvas with too broad of a brush. Inevitably, one public cloud vendor will offer a function driven by a specific, desired AI/ML algorithm while another desirable algorithm driving a different function is only available from a separate public cloud vendor.”

This, says TV Tech, “can erode potential efficiencies and lead to higher costs due to data being shuffled between multiple public cloud services that each charge their own egress and transaction fees.

A sounder approach, it advises, may be to retain control over media assets by storing them in cloud-adjacent capacity and strictly control when and where data may be accessed by any of the hyperscalers for desired AI/ML media processing — retaining ownership and data sovereignty.

4. The Ultimate Business Continuity Solution

This argument gained huge coverage during the pandemic as the companies that had back up in the cloud were able to shift workflows quickly, while others were left stranded and scrabbling to keep up. Cyberattacks are another line of attack for cloud vendors. Put your channel in the cloud and stay on air, they say.

Certainly, data stored by a public cloud vendor off-site is secure — even underground — and safe from most physical threats, deems TV Tech. “Further, the billions of dollars tech giants are spending on cybersecurity is well beyond any sum a broadcaster could hope to match.

But that’s not the entire story. “Media data is vast, continuous and unstructured, not neat and tidy like a spreadsheet. Public cloud backup of the sheer amount of unstructured data a media organization generates daily is a Herculean task. Retrieving it from that storage in a timely way to recover and maintain operations is simply not possible.

An alternative approach, the paper advises, is implementing strategies like permission-based data access and passwords on-site to protect against cyberattacks. Tools outside of the public cloud (hey, Dell again) are said here to offer another level of protection — from AI-monitored log-in and data use activity to air-gapped backup.

5. The Only Logical Migration Solution

Every month TV station groups and other M&E organizations grow nearer to the end-of-life of their racks of standalone SDI-based technology. In doing so, the pressure grows to decide what comes next.

Executives charged with making the decision to refresh hardware are looking for a way out of the capital-spend cycle, and public cloud hyperscalers “appear to offer the perfect alternative where pay-as-you-go pricing is available.”

This paper suggests that again there are other options available to transition operations to OPEX “and accomplish the same goals without sacrificing sovereignty over their company’s data or locking their business into a given public cloud vendor.”

This one is a blatant promotion for Dell, but it’s clear that the CAPEX to OPEX equation is more nuanced than public cloud providers might state.

6. The Cure for High-Priced Real Estate

The final myth outlined here is “The cure for high-priced real estate.” Eliminating hardware, machine rooms, environmental controls and shelves upon shelves of tape, disk and film libraries by relying on compute and storage in the public cloud is an understandable aspiration.

The exorbitant rental cost of maintaining a postproduction in London Soho has been a bone of contention for years. Some facilities have moved to cloud distributed workflows (Jellyfish Pictures, for example) and start-up facilities like Skunk have a business model heavily reliant on the flexibility of cloud.

But not entirely. Skunk just opened a physical HQ in town because its owners (stalwarts of the post production game in the UK) know that clients do still like to visit and sit in a suite with an editor or colorist.

“Even if a TV station wants to downsize its footprint, it is unlikely that it will wish to undo its local presence entirely — even in a high-priced real estate district, such as downtown,” the paper says. “A station is part of the community, and visibility is key. So, jettisoning its studio is likely off the table.”

“Management (at media organizations) must dispel the notion that giving up data sovereignty is simply part of what must be done to realize the benefits of an OPEX-based business strategy. Fundamentally, the workflow should determine the technology choice rather than financial decisions dictating the workflow.”

— “Where Do We Go From Here?”

A station could accomplish much of what it wishes by setting up its own on-premises datacenter — just not in the high-rent part of town — or by deploying infrastructure in a co-lo facility or embracing a multicloud strategy. Connecting its high-visibility location with the compute and storage needed to support newsgathering and production in its own private cloud on the outskirts of town where real estate prices are more affordable achieves the best of both worlds. Stations adopting this strategy also maintain sovereignty over their media data, the paper claims.

These are at very least food for thought and should give companies in the process of strategizing their next infrastructure move a more realistic understanding of the strengths and weaknesses of different vendor arguments.

Other questions hover around the long-term financial consequences of data ingress and egress charges and the comfort level a media organization has with giving up possession and a degree of control over its content.

“Management (at media organizations) must dispel the notion that giving up data sovereignty is simply part of what must be done to realize the benefits of an OPEX-based business strategy. Fundamentally, the workflow should determine the technology choice rather than financial decisions dictating the workflow.”

READ MORE: Where Do We Go From Here? (Dell/TV Tech)

Are you interested in contributing ideas, suggestions or opinions? We’d love to hear from you. Email us here.