Scholar

Media & Entertainment

High-performance Creative Production

Scholar operated on a cloud-only infrastructure hosted in AWS. However, to support changing creative workflows, they needed to deploy high-performance, on-premises workstations for their VFX artist team. Delivering massive, high-resolution media files from the cloud to local workstations traditionally requires bulk data replication, which incurs high bandwidth costs, introduces version conflicts, and causes severe delays that slow production timelines. Scholar needed a way to provide their on-premises team with immediate, local-speed access to their cloud datasets without duplicating data or creating isolated storage silos.

Los Angeles, CA

Scholar is a highly acclaimed creative production and design studio with offices in Los Angeles and New York. They specialize in animation, live-action production, digital content, and visual effects (VFX) for major brands and entertainment networks.

To bridge the gap between their AWS cloud environment and the new physical studio, Scholar deployed a Qumulo Edge Accelerator on-premises. Using Qumulo’s Cloud Data Fabric (CDF), they seamlessly connected the Edge Accelerator to their existing Cloud Native Qumulo (CNQ) cluster in AWS. This created a hub-and-spoke architecture in which CNQ in AWS serves as the authoritative, scalable data core, while the Edge Accelerator functioned as an intelligent caching endpoint for local workstations.

• Local-Like Performance for Cloud Data: The Edge Accelerator uses Qumulo’s AI-driven NeuralCache to anticipate the creative team's access patterns and intelligently prefetch required media files from AWS into local NVMe storage before the VFX artists even request them. This provides the on-premises workstations with sub-millisecond, local-like performance, overcoming WAN latency without requiring the entire dataset to be downloaded. This also allowed them to use higher-performance local GPUs for render farming. 

• Zero-Copy Global Collaboration: Cloud Data Fabric projects a single, globally visible namespace across both AWS and the on-premises studio. Scholar's team can access and collaborate on the same live media assets simultaneously without ever creating duplicate copies, entirely eliminating data sprawl and version reconciliation issues.

• Real-Time Edge-to-Cloud Synchronization: When VFX artists create or modify files on their local workstations, write operations are instantly committed to mirrored local SSDs on the Edge Accelerator, minimizing latency. These changed data blocks are then continuously synchronized back to the CNQ hub in AWS. This ensures that any cloud-based services Scholar uses, such as AWS burst rendering farms, always interact with the most current version of the project.

• Reduced Infrastructure Footprint: By projecting data on demand rather than replicating it, Scholar minimized the physical storage hardware needed at their on-premises studio, controlling costs while maintaining complete flexibility.