Implementing Hyperscale and HPC Techniques and Technologies Necessary to Modern Media and Broadcast Workflows

摘要

Across the media production and broadcast industries, studios of every size are accelerating their consumption of vast amounts of storage resources. With increasing quantities of data produced from increasingly higher-resolution digital cameras and content featuring greater color depth across a broader color spectrum, file manipulation and storage capacity have now reached the scale and performance requirements of high-performance computing (HPC) applications. Broadcast, production, post-production and distribution now generate hyperscale workloads with tens to hundreds of petabytes of business-critical data and streams of several gigabytes/second (GB/s) per client. For media companies to remain competitive, they need to apply the lessons learned by hyperscaler and HPC companies to cost-effectively and consistently provide high levels of performance with a very small team while ensuring long-term data protection. — This paper and presentation will utilize customer and industry examples from hyperscaler, HPC, and Media and Entertainment (M&E) organizations to discuss in tutorial fashion and learn how to implement a hyperscaler-like storage infrastructure that provides limitless scalable storage to all applications across a broad range of workloads; utilize commodity servers without hardware redundancy, RAID controllers or NVRAM required for the lowest cost and flexibility; and remove the challenges of deploying and managing infrastructure, enabling new methods of creating and delivering content.