David Sallak is the VP/CTO, M&E for Sunnyvale, CA’s Panasas (panasas.com), the provider of high-performance storage solutions. Here, he looks at how studios might employ cloud storage to make the most use of their assets.
Taking stock of change in the production of media & entertainment content in 2017 leads to recognizing that while some things stay the same, much has evolved and some entirely new concepts are about to shake things up. The business of media content creation relies on some foundational basics such as file-based workflows, resulting in the need for more of everything — compute and graphics power, bandwidth and capacity. What’s changed is the way we can interact with the content we produce.
New insights have always been desired from the content we make, and there’s a sense that unrecognized value exists within the historical archives of our media vaults. Many attempts have been made to apply asset management values to content libraries, but a barrier to widespread leverage is the complexity of tagging assets with useful metadata by which everything can be reviewed at scale, in a manner that can be trusted to have been tagged accurately. In line with this need is awareness that media assets change their value based on constantly emerging relationships. This is due to correlation variation by which tagged assets can mutate in value over time.
How can this potential for greater media value be achieved without overwhelming our budgets and capabilities? It turns out this is a perfect use case for the cloud. The elastic scalability of search algorithms in the cloud has one very compelling value to offer all media producers, and that’s the emerging space of artificial intelligence and machine learning. By exposing assets to secure cloud compute, media companies will see the emergence of the “Easy Button” of metadata and media asset management coming together.
Media production also needs greater agility and flexibility in infrastructure. Our industry is ripe for change in how technology is utilized, which leads to insights on the role of software bringing together compute and storage in new and powerful ways. Human-driven tasks require technology to perform at its best, in order to maintain the magic of imagination and creativity for artists and editors. Many secondary tasks don’t require people to perform them, and these are areas where non-realtime tasks can leverage new trends in technology to offer better ways to get secondary work done.
To tap these new trends, the technology buyer should look for industry-standard platforms that support concepts of virtualization and containers, which allow many different application stacks running these media-based tasks in the background in a steady-state fashion. Most of these tasks don’t require constant infrastructure, which leads to scheduling and configuring much of the facility to respond to API triggers and get the best utilization from optimized hardware resources. Storage in particular should support hosting utility-based software tasks against media assets, in a scalable fashion.