VDURA

What VDURA does

VDURA (formerly Panasas, rebranded May 2024) is a privately held California-incorporated data infrastructure software company that builds the VDURA Data Platform, a software-defined, shared-nothing scale-out architecture that combines a true parallel file system (PanFS, originating in 2001) with object storage under a single unified namespace. The platform supports up to 1,500 nodes per namespace with linear performance scaling, mixes NVMe flash and HDD capacity tiers via Dynamic Data Acceleration, and is anchored by the proprietary VeLO flash-optimized metadata engine capable of up to 100 million IOPS. VDURA sells integrated V5000 all-NVMe flash and V7000 hybrid hardware systems alongside the software platform, and supports software-defined deployment on certified commodity hardware. Revenue is generated through multi-year software subscriptions (including 24/7/365 support), hardware system sales, professional services, and training.

VDURA serves AI factories, neoclouds, federal agencies, and global enterprises running GPU-accelerated AI training and inference, traditional HPC workloads, and mixed HPC/AI environments. Named deployments span defense and aerospace (Raytheon, Northrop Grumman, Boeing, Pratt & Whitney, a large federal system integrator at ~200PB), federal civilian agencies (NIH, Federal Reserve), energy (a global energy leader at ~26 PiB unified capacity, PGS), life sciences and academia (Hormel Institute, Vanderbilt 20,000-core cluster, Goethe University, Santa Clara University), and technology (NIO). Go-to-market combines direct enterprise field sales with technical briefings and proof-of-concept deployments and a channel-first partner program that includes technology alliances with AMD, NVIDIA, Penguin Solutions, Western Digital, Seagate, Phison, NEC, Cornelis Networks, and resellers such as BIOS IT, SourceCode, and Starfish Solutions. The company's strategic positioning emphasizes mixed-fleet economics — claiming up to 60% lower TCO versus all-flash competitors via the Flash Relief Program — as a counter to flash price volatility in AI infrastructure builds.

Name

VDURA

Legal name

VDURA, Inc.

Website

https://vdura.com

Company type

Private

Founded year

2001

Operating status

Operating

Headcount range

51–100 employees

Short description

VDURA (formerly Panasas) provides software-defined AI and HPC data storage combining parallel file system performance with object storage economics. It serves AI factories, neoclouds, federal agencies, and enterprises across research, energy, defense, and life sciences.

Ownership category

akta.pro rank

Product category

AI & HPC Data Storage Infrastructure

NAICS

Software Publishers (5132), Computer Storage Device Manufacturing (334112)

SIC

Computer Storage Devices (3572)

akta.pro primary industry

Virtual Desktop & Application Delivery (VDI/DaaS, Remote Apps) (BPAEAIAC)

Keywords

• Parallel file system
• AI data storage
• HPC storage
• Data infrastructure
• Flash storage systems

Headquarters

HQ city

Milpitas

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Operations, Marketing or Sales, Supply Chain, Infrastructure

Revenue model

1. Software Platform License and Subscription: VDURA Data Platform software is sold as a subscription including software licenses, support (24/7/365), and updates. The platform is software-defined and can run on certified hardware platforms.
2. Hardware Systems (V5000, V7000): VDURA sells integrated hardware systems including the all-NVMe V5000 flash system and hybrid capacity configurations. Hardware includes certified platforms ensuring tight integration and optimal performance.
3. Professional Services: Customized strategies and implementation services delivered 24/7/365 to meet specific customer business needs. Includes deployment, migration, and optimization services.
4. Training Courses: Technical training and enablement services for customer teams to build expertise with VDURA platforms.

Pricing tiers

Go-to-market motion2 records

Distribution channels3 records

Marketing channels10 records

Core offering

VDURA develops and sells the VDURA Data Platform, a software-defined parallel file system that combines NVMe flash performance with HDD cost-efficiency in a unified namespace for AI and HPC workloads. The company offers the V5000 All-NVMe Flash System and V7000 hybrid platform as integrated hardware systems running the data platform software, with features including the VeLO metadata engine, vPods scale-out storage units, and DirectFlow parallel access clients.

Product overview

VDURA offers a unified data infrastructure platform combining the VDURA Data Platform software with V5000 and V7000 hardware systems. The V5000 is an all-NVMe flash system delivering 7.2M IOPS and 210 GB/s throughput, while the V7000 platform adds RDMA support for enhanced AI workloads. Both platforms run the VDURA Data Platform software featuring the VeLO metadata engine for flash-optimized performance, vPods parallel storage units, DirectFlow POSIX-compliant clients, and intelligent tiering across NVMe and HDD media. The company supports AI and HPC workloads through tools like the GPU Storage Calculator and Flash Relief Program, with products designed for petabyte-scale capacity requirements across AI factories, neoclouds, and enterprise deployments.

Differentiator

Problem solved

Functional benefit

Brands

• V5000: All-NVMe Flash System platform delivering up to 7.2M IOPS and 210 GB/s throughput, designed for AI and HPC workloads with VeLO metadata engine.
• V12 Data Platform
• VeLO
• PanFS
• Flash Relief Program
• ActiveStor Ultra

Products and services

• VDURA Data Platform Software-defined data infrastructure platform combining true parallel file system performance with resilient object storage under one control plane. Supports V5000 and V7000 hardware platforms, delivering flash-optimized architecture with intelligent tiering across NVMe flash and HDD tiers for AI factories, neoclouds, federal agencies, and large enterprises.
• VDURA V5000 All-NVMe Flash System All-NVMe flash storage hardware system delivering 7.2M IOPS and 210 GB/s throughput with effective capacity starting at 286TB. Powered by the VeLO metadata engine for ultra-low latency, designed for 24x7x365 AI workloads with intelligent automation. Includes RDMA support added in 2026.
• VDURA V7000 Platform Enterprise storage platform with RDMA support designed for large-scale AI workloads with enhanced GPU utilization and storage efficiency. Runs the VDURA Data Platform software.
• VDURA Professional Services Customized strategies and implementation services delivered 24/7/365 to meet specific customer business needs, including deployment, migration, and optimization services.
• VDURA Training Courses Technical training and enablement services for customer teams to build expertise with VDURA platforms.

Quantifiable outcome

• Up to 20x improvement in metadata operations
• +8 more outcomes

Named customers14 records

Segments9 records

Ideal customer profiles7 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration1 record

AI capability3 records

Feature12 records

Partnerships

Eight partnerships are on record, tiered core and minor.

• Western DigitalcoreTechnology or Integration · 10 March 2026Strategic collaboration to certify WD's Ultrastar Data60 and Ultrastar Data102 Hybrid Storage Platforms within VDURA Data Platform architecture. Creates unified hybrid-tiered storage solution pairing VDURA's all-NVMe V5000 flash tier with WD's Ultrastar Hybrid Storage capacity tier, managed within single namespace for petabyte-scale requirements. Aims to deliver hyperscale storage economics to enterprise AI infrastructure.
• PhisoncoreTechnology or Integration · 19 November 2025VDURA and Phison shattered AI data transfer record at SC25, transferring 136.5 petabytes in 27 seconds with Hafthor Bjornsson. Event showcased combined power of Phison's SSDs and VDURA's data infrastructure for AI and HPC environments. Phison's Pascari SSDs used in world record demonstration.
• SourceCodecoreStrategic or Co-development Partner · 11 July 2025SourceCode launched Atlas AI Ignite solution powered by VDURA Data Platform, designed for rapid AI deployment with quick time-to-value. Atlas AI Cluster delivers maximum performance. VDURA featured in SourceCode webinar on scaling AI workloads with hybrid cloud and on-prem solutions.
• AMD and Penguin SolutionscoreStrategic or Co-development Partner · 21 February 2025Collaboration involving VDURA, AMD, and Penguin Solutions to provide HPC and AI infrastructure for energy industry. Companies offer integrated solutions that simplify data workflows, increase processing speed, and enhance sustainability. VDURA partnered with AMD to introduce industry's first scalable GPU reference architecture with AMD Instinct accelerators.
• Channel Partners (General)coreChannel Partner/ Reseller/ DistributorVDURA operates channel-first partner program with deal registration, partner certifications, and co-marketing initiatives. Focus areas include AI integration for partner enablement, expanded certification programs, deeper MSP and SI engagement, and profitability improvements.
• Penguin SolutionscoreStrategic or Co-development PartnerPenguin Solutions' OriginAI scalable, pre-validated AI factory infrastructure leverages VDURA storage to enable GPU clusters from 48 to 16,000 GPUs. Results: faster time-to-value, predictable performance, and secure enterprise-grade multitenancy.
• BIOS ITminorChannel Partner/ Reseller/ DistributorBIOS IT partnership covering life sciences, genomics, and finance deployments. BIOS IT serves as a value-added reseller and system integrator for VDURA solutions.
• RadiumminorStrategic or Co-development PartnerVDURA + Radium partnership to build private AI with enterprise-grade performance. Joint go-to-market for private AI infrastructure solutions.

Scale indicators15 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Vast Data: AI-native data platform combining a parallel file system, object storage, and database under one namespace. Competes head-to-head with VDURA in AI factory, neocloud, and large enterprise deployments, with a similar software-defined, scale-out architecture and overlapping customer target list.
• Weka: AI and HPC data platform built on a parallel file system optimized for GPU workloads. Targets the same AI factory, neocloud, and enterprise AI customer base as VDURA with comparable claims of NVMe-class performance, RDMA, and GPU utilization advantages.
• DDN (DataDirect Networks): Long-established HPC and AI storage vendor with parallel file system products (EXAScaler, IntelliFlash). Directly competes with VDURA in federal, research lab, and large-scale HPC/AI deployments with overlapping customer segments in defense, life sciences, and academia.
• Hammerspace: Parallel file system and global data orchestration platform for AI and HPC. Targets similar AI training and enterprise HPC customers with a software-defined approach to unifying file and object data across on-premises and cloud.

Emerging players

• MinIO: Open-source, S3-compatible object storage software often deployed for AI/ML data lakes. Competes in the AI data infrastructure stack with VDURA, particularly in Kubernetes-native and cloud-native AI pipelines, but lacks a parallel file system and integrated hardware story.
• Qumulo: Software-defined file and object storage platform for AI, HPC, and unstructured data workloads. Competes with VDURA in AI data infrastructure and enterprise HPC with a hybrid cloud and multi-protocol approach, though smaller and less HPC-parallel-file-system-focused.
• Cloudian: S3-compatible object storage platform used for AI/ML data lakes, backup, and large unstructured data sets. Overlaps with VDURA in the AI data infrastructure layer but focuses on object storage rather than a parallel file system for HPC and AI training workloads.

Broad incumbents

• Pure Storage (FlashBlade): Public incumbent in all-flash enterprise storage; FlashBlade product line is a direct competitor in AI/HPC unstructured data workloads. Operates a much broader portfolio (block, file, object, SaaS) and competes with VDURA on performance and scale across enterprise AI deployments.
• NetApp: Broad enterprise storage incumbent with file (ONTAP), object (StorageGRID), and AI-optimized products. Competes with VDURA in enterprise AI, federal, and HPC segments, though its core product set is more general-purpose and not parallel-file-system-first.
• IBM Spectrum Scale (GPFS): Long-standing HPC parallel file system and enterprise storage platform bundled with IBM hardware, software, and cloud. Overlaps with VDURA's HPC, federal, and large enterprise customers with a much broader, integrated portfolio.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

• Xx.com/vdurainc
• YouTubeyoutube.com/@VDURAINC
• LinkedInlinkedin.com/company/vdurainc
• Instagraminstagram.com/vdurainc
• Facebookfacebook.com/profile.php

Revenue estimate

Valuation estimate

Number of profiles

Profiles14 records

Funding overview

Funding rounds9 records

Investors9 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about VDURA