The Lustre Collective Brings Independent Expertise to High-Performance Storage Market

The Lustre Collective launches independent support for parallel filesystem serving exascale HPC and AI infrastructure across clouds and data centers.

Presented at the 66th IT Press Tour, January 2026

Three veteran filesystem developers have launched The Lustre Collective, creating an independent support and development company for organizations running large-scale AI and high-performance computing workloads.

The company addresses a specific market need. Lustre powers critical infrastructure for government labs, hyperscalers, and AI companies, but support options have historically been tied to specific hardware vendors. TLC offers vendor-neutral expertise for the parallel filesystem that handles some of the most demanding storage workloads in production today.

Market Position and Customer Base

Lustre runs on the storage systems behind major exascale supercomputers. The Frontier supercomputer at Oak Ridge National Laboratory uses Lustre. So does El Capitan at Lawrence Livermore National Laboratory. These systems operate at scales that few other filesystems can handle.

The technology has also moved into commercial AI infrastructure. NVIDIA's EOS AI DGX SuperPOD relies on Lustre. xAI's Colossus AI supercomputer does too. The filesystem handles the massive data throughput requirements these systems demand.

Eight of the top 10 HPC systems and over 60% of the top 100 systems use Lustre. That market penetration reflects 25+ years of real-world deployment and refinement.

Every major public cloud now offers Lustre as a first-party service. AWS, Microsoft Azure, Google Cloud, and Oracle all provide managed Lustre options. This availability across cloud platforms gives enterprises flexibility in where they run their HPC and AI workloads.

The Team and Their Credentials

The founders bring deep technical expertise and business experience to the company.

Andreas Dilger serves as CTO. He has led Lustre development since 1999 and authored key papers on the filesystem. His involvement in Linux and open-source communities spans decades. He reviews papers for major technical conferences.

Peter Jones is CEO. He has managed large Lustre engineering teams at multiple organizations and holds leadership positions in the Lustre community, including work with EOFS and OpenSFS.

Colin Faber leads engineering as Director of Engineering. He was involved with Lustre from early days and has built several well-known Lustre storage appliances. He advocates for open-source approaches and has been active in the community for years.

This combination of technical depth and business leadership matters. The team knows the technology and understands how to support enterprise deployments.

Service Offerings

TLC provides several types of engagement tailored to customer needs:

• Expert consulting and training help organizations optimize their deployments and train internal teams. This includes performance tuning and optimization work.

• Production support contracts give organizations access to the core development team when issues arise. Response times and support levels vary based on subscription tier.

• Custom feature development lets organizations sponsor specific capabilities they need. The code gets contributed back to the open-source project, benefiting the entire community.

• Deployment and migration services help organizations move to Lustre or upgrade existing systems. The team brings experience from hundreds of deployments.

Technology Roadmap

The 2026 roadmap includes features that address enterprise requirements:

• Erasure-coded files reduce storage overhead while maintaining data protection. This is important for organizations managing petabytes of data where storage efficiency directly impacts budget.

• Trash can and undelete capabilities provide recovery options when users accidentally delete files. This is basic functionality in consumer systems but has been missing from Lustre.

• Fault-tolerant management services improve system reliability. The Management Service is critical to Lustre operations, and redundancy reduces the risk of outages.

• Client-side data compression reduces storage requirements and network bandwidth consumption. Both affect operational costs.

Strategic Investment Areas

TLC plans to expand its team in 2026 to focus on six areas:

1. Improved availability mechanisms reduce downtime and make maintenance easier.

2. Better serviceability and ease of use lower the total cost of ownership.

3. Enhanced multi-tenancy capabilities let organizations safely share infrastructure across teams or customers.

4. Easier quality of service management gives administrators better control over resource allocation.

5. Modernized tooling and monitoring help teams manage complex deployments.

6. Expanded filesystem-level resiliency protects against data loss and system failures.

These priorities reflect what enterprises need as they scale up AI and HPC workloads.

Why Independent Support Matters

The open-source nature of Lustre means multiple companies contribute code. Oracle, DDN, AWS, Google, Microsoft, HPE, NVIDIA, and national laboratories all commit changes to recent releases.

This creates both opportunity and complexity. Organizations benefit from broad community investment, but they need partners who can work across the entire ecosystem rather than pushing a single vendor's roadmap.

TLC positions itself as that vendor-neutral partner. The company can recommend the best approach for a given deployment without being tied to specific hardware or cloud platforms.

Business Impact

For organizations running large-scale AI training or HPC workloads, storage performance directly affects compute ROI. GPUs and CPUs sit idle when waiting for data. Lustre's ability to deliver 10TB/s+ throughput and 100M+ IOPS helps keep expensive compute resources productive.

The filesystem's scalability matters too. Supporting 20,000+ clients and 100,000+ GPUs in a single namespace simplifies management compared to stitching together multiple smaller systems.

Security and isolation features let organizations safely run multi-tenant environments. Encryption, authentication, and nodemap controls meet enterprise requirements for data protection and access control.

Looking Ahead

TLC will announce a detailed long-term roadmap at LUG2026 in April. The company is consulting with partners to identify which projects will deliver the most value.

For organizations already running Lustre, TLC offers an independent option for support and development. For those evaluating storage options for AI or HPC workloads, the company provides expertise on whether Lustre fits their requirements and how to deploy it effectively.

The company's website is thelustrecollective.com.