Choose your cluster
Which cluster is right for your research?
Compare at a glance, then dive into each cluster's page for specs, queues, and access details.
| Cluster | Best for | Scale | PHI data | Who can use it |
|---|---|---|---|---|
 SeaWulfThe general-purpose workhorse |
Physics & climate science | 400+ nodes · 23,000 cores | No PHI | Campus community & Brookhaven Lab staff |
 NVwulfGPU acceleration for AI & ML |
AI & machine learning | 24× H200 NVL · 80 PFLOPS FP8 | No PHI | Campus community |
 ClinWulfHIPAA-compliant computing on PHI |
Clinical research on PHI data | 30 nodes · 32 GPUs · 1.5 PB | HIPAA ✓ | Medicine Campus |
 SecureWulfSecure computing for controlled data |
Controlled-access and regulated research data (NIST SP 800-171) | 8 GPUs · 128 CPU cores (planned) | No, PHI stays on ClinWulf | Coming Christmas 2026 |
Regulated & controlled data
Computing for controlled research data
A NIST 800-171-aligned secure enclave, built on TiCrypt, in support of NIH and biomedical datasets that carry controlled-access requirements, including dbGaP controlled-access genomic datasets, the All of Us Research Program controlled tier, NCATS Clinical and Translational Science Awards, and NCI Cancer Moonshot data-intensive grants. Learn about SecureWulf or contact RCI to plan ahead for projects that will need it.
🛡️ CMMC & CUI research
For research regulated under CMMC or involving Controlled Unclassified Information (CUI), Stony Brook University utilizes Sherlock Cloud — a managed, compliant environment for CUI and CMMC-regulated research computing and data. Contact RCI to get started.
Everything else
Docs, software, and support
📚 Documentation
One task-based library — connecting, jobs, storage, software — tagged by cluster.
📦 Software
One searchable catalog with per-cluster availability.
🧾 Job script builder
Fill in a form, get a ready-to-submit sbatch script for any cluster.
🖥️ Open OnDemand
Jupyter, RStudio, MATLAB, and more — no SSH required.
🛟 Support
Tickets, virtual office hours, and the top questions answered.
🧑💻 Student assistants
The students who help run our clusters — and where the program takes them.
News
HPC announcements
July 7, 2026
MaintenanceNVwulf Cluster Maintenance
The NVwulf cluster will undergo maintenance starting at 10 AM on Tuesday, July 28th, concluding by end of business that day. All login nodes, compute …
June 26, 2026
MaintenanceScheduled Data Center Maintenance
Multiple GPU and core queues go offline at 4 PM on Monday, July 13th for electrical work, expected to conclude by lunch on July 14th. Affected: 28-cor…
June 2, 2026
MaintenanceSeaWulf Intel Sapphire Rapids Maintenance
HBM queues and the xeonmax login node go offline at 5 PM on Monday, June 15th for electrical maintenance on data center circuits. Systems should retur…