VDURA adds multi-tenant control plane for AI storage

VDURA has introduced a new multi-tenant control plane and updated S3 features for AI and high-performance computing environments. The additions are planned for general availability in the second half of 2026 for V5000 class systems.

The update includes three changes: a redesigned control plane for storage management, higher S3 throughput for cloud-native AI workloads, and native S3 object tagging. Existing customers will be able to deploy the software through an in-place online upgrade.

The new control plane is aimed at teams managing shared storage across multiple users or groups. It adds a new management interface and a tenant administration model designed to simplify oversight of complex storage estates.

VDURA is also adding a REST API for key platform operations to support automation and integration for operators managing storage through internal tools and workflows.

Multi-tenant administration has become a pressing issue for organisations running AI and HPC systems, where large user bases often share the same infrastructure. Storage teams must balance access, governance, and operational control while handling fast-growing data volumes from model training, inference, and scientific computing.

S3 changes

A second part of the announcement focuses on S3 performance. VDURA has made targeted improvements intended to sustain throughput across cloud-native AI pipelines, including model checkpointing, inference serving, and large-scale dataset ingestion.

The changes reduce latency for S3-native operations and increase aggregate throughput for concurrent reads and writes. That matters for AI operators that depend on object storage during training runs and production inference, especially when multiple jobs access data at the same time.

VDURA is also introducing native S3 object tagging. The feature lets organisations attach metadata to stored objects so they can apply policies and controls across large data collections.

Object tagging is commonly used to organise data, define retention rules, and manage access rights. In AI and research settings, it can also help distinguish training datasets, model artefacts, and other files as they move through different stages of use and governance.

The feature supports policy-based lifecycle management, automated tiering workflows, and more granular access controls. For operators of large object stores, metadata tools are becoming increasingly important as data estates expand and compliance demands become harder to manage manually.

Operational focus

The package reflects a broader industry shift in AI infrastructure from raw hardware build-out to operational management. As storage systems scale across departments and tenants, suppliers are placing more emphasis on software controls, interfaces, and interoperability with cloud-native tools.

VDURA framed the latest changes around those operational concerns rather than new hardware. The control plane and S3 updates are intended to ease day-to-day administration while supporting data-intensive pipelines.

Chris Girard, Vice President of Product Management at VDURA, described the release as a response to customer demand for easier management and stronger data workflow support.

"With these new capabilities, we're focused on making VDURA more powerful and easier to operate at every level of the organization," said Chris Girard, Vice President of Product Management at VDURA. "The new management interface brings clarity and control to the people running VDURA day-to-day, while our expanded S3 capabilities give data engineers and AI practitioners the tools to build more sophisticated, automated data pipelines. These are the operational and integration capabilities our customers have been asking for."

The rollout is planned for all V5000 class systems in the second half of 2026. Existing customers can upgrade in place via an online software update.