VDURA V12 boosts AI data speed, cuts storage costs by over 20%
VDURA has introduced Version 12 of its Data Platform, incorporating new features designed to boost scalability, reduce storage costs, and enhance resilience for artificial intelligence (AI) and high-performance computing (HPC) workloads.
Performance enhancements
The updated platform features an Elastic Metadata Engine designed to support rapid scaling across distributed units. This new engine enables metadata operations to run linearly across multiple nodes, supporting billions of files and objects while eliminating traditional performance bottlenecks. According to the company, this advancement can accelerate metadata operations by up to 20 times compared to previous versions.
VDURA also reports a more than 20% increase in aggregate throughput performance with V12. The upgrade is tailored to handle the demanding requirements of AI model training, inference workloads, and movement of hyperscale datasets. These performance improvements are accompanied by a claimed reduction of over 20% in cost per terabyte, further enhancing the cost efficiency of storage-intensive operations.
"With VDURA V12, we're increasing throughput performance by more than 20%, accelerating metadata operations by up to 20X, and reducing cost per terabyte by over 20%," said Ken Claffey, CEO, VDURA. "At the same time, we're simplifying data protection with new snapshot capabilities that make backup and recovery effortless at any scale."
Data protection
The update introduces system-wide Snapshot Support, enabling instantaneous, space-efficient point-in-time copies of datasets. This capability is targeted at AI pipelines, enabling users to create model checkpoints and enhance operational recovery. Snapshots are compatible with both manual and policy-based retention, streamlining backup and disaster recovery options for large-scale environments.
SMR HDD integration
V12 marks the introduction of support for Shingled Magnetic Recording (SMR) hard disk drives, which are designed for ultra-dense data storage. The platform's new write-placement engine organises sequential zones across SMR drives, unlocking an estimated 25-30% more capacity per rack. This effort aims to address growing data volumes while maintaining the throughput and integrity of stored information.
Sector context
The new release arrives amid increasing demand from large enterprises, AI service providers, and emerging cloud operators for both scale and efficiency. V12's enhancements are designed to help organisations manage increasingly large datasets and workflows while maintaining control over costs and resilience.
Mary Jander, Senior Analyst at Futuriom, commented on the introduction of the new version. "With its V12 release, VDURA expands its highly efficient parallel data management architecture with distributed, elastic metadata scaling; automated snapshotting; and SMR HDD optimisation."
"With these enhancements, VDURA is positioned to meet the growing demands of large enterprises, AI service providers, and neoclouds for optimal performance in increasingly complex data environments-at scale and without compromising cost efficiency or resilience."
