Hammerspace for Higher Education & Research
Research Computing Is Evolving from HPC to HPC + AI
Universities and research institutions are adding GPU infrastructure and AI workloads alongside traditional modeling, simulation, and scientific computing. At the same time, decades of valuable research data remain distributed across generations of parallel file systems, NAS-based home directories, object storage, tape archives, and other infrastructure.
Research computing organizations need to modernize for HPC + AI while preserving decades of research data, making the most of existing infrastructure, and avoiding the cost of putting every petabyte on expensive flash storage.
“Hammerspace…allows for immense flexibility, immense cost savings, and covers all cluster workload profiles.”
Exec Director, Research Computing Operations, ACCRE
Vanderbilt Advanced Computing Center
Unify Research Data Across HPC and AI
A global namespace gives researchers a standards-based way to access data across storage platforms, tiers, sites, and clouds without needing proprietary client software. Data-in-Place Assimilation brings existing research data into that environment without time-consuming data migrations. Automated data orchestration then non-disruptively places data where it needs to be across any on-prem or cloud-based storage according to policies for performance, capacity, and cost.
Hammerspace's standards-based parallel file system delivers the performance required by demanding HPC and GPU workloads without proprietary client software or hardware lock-in.
Unified Data Plane
Automated Data Services
Accelerated Data Orchestration
Silo-less Data Storage
Key Benefits
One Standards-Based Data Platform for HPC + AI
Run traditional HPC and emerging AI workloads on one high-performance data platform.
Unlock Decades of Research Data
Make existing research data available to new HPC and AI infrastructure without disruptive migrations.
Keep GPUs and CPUs Productive
Reduce time spent waiting for data so GPU and CPU resources can spend more time on research workloads.
Modernize Infrastructure and Reduce Costs
Use existing and new storage together to optimize performance, capacity, and cost.
“Normally with an equivalent system… we have at least three different file systems in play to keep things going,” says the Director of the university High Performance Computing Center. “We needed something that we could deploy that would serve a wide variety of needs… with the minimum complication.”
See how a large public research university modernized their infrastructure for AI with Hammerspace
Read the Case StudyCommon Use Cases
Scratch & Simulation
Use server-local NVMe as a shared high-performance storage tier for GPU and CPU research workloads. Example uses include:
- Checkpointing and restart
- Hot training and inference datasets
- Data-intensive simulation workloads
Project & Shared Research Workspaces
Persistent, high-performance storage for the bulk of HPC data:
- Shared team workspaces for research groups and institutes
- Multi-tenant AI and data science environments
- Cross-vendor storage pools unified under one namespace
Home Directories, Collaboration & Visualization
Enterprise-grade data services for user environments:
- Researcher home directories
- Departmental shares and visualization pipelines
- Mixed protocol access (NFS, SMB, S3) with snapshots and quotas
Archive & AI Reuse of Historical Data
Keep archived research data visible and available for reuse through supported archive integrations:
- Global visibility into archived datasets
- Recall archived files when needed and use policies to manage placement across storage tiers
- Access to historical datasets for reuse in HPC and AI workflows