September 13th, 2018 by Adam Armstrong
Qumulo Ups Performance & Lowers Costs For M&E
Today Qumulo Inc. announced that it was making several software innovations that will increase performance in its file storage solutions. These innovations leverage machine learning capabilities with the specific goal of transforming the media and entertainment (M&E) industry through higher performance and lower economics for tasks such as video editing, visual effects and video distribution. The new software innovations will be implemented across all Qumulo product lines.
Video technology is improving at a rapid rate. Video quality and visual effects have made tremendous leaps and bounds recently and legacy storage equipment has not been able to keep up. Companies can’t upgrade hardware every time these technologies improve but they can upgrade software. According to Qumulo, its new innovations can help studios eliminate SANs, client-side software, HBAs, and fibrechannel switches and cabling through its high-performance centralized uncompressed 4K and 6K video playback solutions. Instead studios will be enabled through 100% Ethernet environments with all-flash storage. The increase in performance means more editors can collaborate on a project at once, potentially decreasing its time to finish and lowering costs.
The performance improvements include:
- Qumulo P-Series
- Highest all flash performance per TB in the industry. The P-Series sets a new standard: a 96TB cluster offers 18GB per second throughput, featuring 200,000 input / output operations per second (IOPS) from a four-node cluster. This means 90GB/sec/rack in a single unified namespace with instantaneous quotas and per-directory snapshots. As the minimum configuration, capacity and performance scale from here.
- All flash performance on a hybrid platform. The latest software offers the industry’s first ever scale-out NAS support of centralized uncompressed 4K and 6K video playback using machine learning and state-of-the-art NVMe drives.
- Qumulo File Fabric (QF2)
- Pre-fetch Intelligence. Machine learning-powered pre-fetch anticipates and queues up the files that users will most likely access next, whether it be a sequence of files by name, or simply a traversal of all files in a directory structure.
- Proactive Caching. File access pattern analysis proactively caches individual blocks of files most likely to be repeatedly accessed.
- Maximum flexibility and investment protection. All platforms come with native replication directly to the public cloud.
The performance improvements are available now to all existing Qumulo subscriptions and all existing QC104, QC208 and QC260 systems for no additional charge.