ClinWulf is where RCI's clusters handle protected health information. It's for IRB-approved studies in Health Sciences and Medicine, and keeps notebooks, R sessions and viewers on ClinWulf's own nodes and storage.
Get connected
Open OnDemand
Jupyter, RStudio, VS Code, a Linux desktop, IGV, ParaView and TensorBoard, running inside ClinWulf.
Network access: only from within Health Sciences and Stony Brook University Hospital and Medical Center. From anywhere else, connect through Citrix Desktop or the Stony Brook Hospital & Medical Center VPN first.
Launch Open OnDemandSSH
The ClinWulf login server. Home folders are at /vast/home/<netid>.
ssh <netid>@login02.uhmc.sbuh.stonybrook.eduSpecifications
| Resource | Details |
|---|---|
| Compute nodes | 30 |
| GPUs | 24, including 8 NVIDIA A100 |
| High-memory nodes | 11 nodes, 750 to 1,000 GB RAM |
| Storage | 1.5+ PB VAST |
| Network | High-bandwidth Ethernet and InfiniBand |
Queues
| Queue | Longest session | Use it for |
|---|---|---|
| CPU short | Short sessions | Quick interactive work |
| CPU long | 8 hours | A working day of analysis |
| CPU extra | 10 days | Long-running jobs |
| GPU short | 2 hours | Quick GPU tests |
| GPU long / extra | 8 hours / 10 days | Deep learning and imaging |
Pick a CPU queue unless your work uses a GPU. Each launch form shows the limit for every queue.
Apps in your browser
Best suited for
- Clinical research on PHI
- Medical imaging at scale
- Genomics with identifiable data
- HIPAA-regulated machine learning
Getting access
- Have an IRB-approved studyClinWulf is available to IRB-approved studies in Health Sciences and Medicine working with PHI.
- Contact RCIOpen a support ticket and mention ClinWulf, your study and who needs access.
- Set up Duo, then sign inUse the ClinWulf OnDemand portal or the login server. From outside Health Sciences and Stony Brook University Hospital and Medical Center, connect through Citrix Desktop or the Stony Brook Hospital & Medical Center VPN first.
Related guides
Also available

The general-purpose workhorse: CPU, high-memory and A100 GPU nodes.
View cluster โ
GPU acceleration for AI and machine learning, on H200 and Blackwell GPUs.
View cluster โ
Secure computing for controlled-access research data.
View cluster โ