Vdura: AI Storage Cost Control & Durability
“`html
vdura: From Panasas to Software-Defined Data Infrastructure for AI and HPC
Table of Contents
What is Vdura?
Vdura provides a software-defined data infrastructure platform built for Artificial Intelligence (AI) and high-performance computing (HPC). Their software combines the scalable, linear performance of a parallel file system with the resilience and cost efficiency of object storage, providing a unified global name space with one control plane and one data plane.
The Rebranding: Panasas to vdura
Panasas rebranded itself as Vdura on May 7, 2024, marking a strategic shift from selling proprietary hardware to focusing on software-defined storage with a subscription-based business model for AI and high-performance computing. This transition allows Vdura to cater to a broader market and offer more flexible deployment options.
A History of Innovation: From PanFS to Vdura
panasas was founded in 2000 by Garth Gibson,a pioneer in storage technology – co-inventor of RAID and the creator of the first Linux-based parallel file system,PanFS. In 2004, Panasas began producing high-performance storage systems based on PanFS. The core technology of PanFS has evolved into the foundation of Vdura’s software subscription offerings. Notably, Garth gibson returned to vdura as chief Technology and AI officer in September 2025, bringing his extensive expertise back to the company.
Addressing the Demands of AI Workloads
Supporting AI workloads requires sustained throughput at scale to prevent GPU starvation and maximize investment. AI also demands expanding data metadata capacity and performance for rapid data access. Vdura claims to accelerate all parts of the AI pipeline - data ingesting, model loading, training, checkpointing, fine-tuning, and inference – unlike some competitors who focus on specific stages.
Various elements of the AI data pipeline
Vdura
The Vdura Data Platform (VDP)
The company states that their Vdura Data Platform (VDP) can transform thousands of storage servers into a high-performance, resilient, and durable data platform. It is indeed designed to withstand numerous failures of both devices and nodes, is straightforward to deploy and manage, and seamlessly integrates NV