National Research Platform: Nautilus Hyperconverged GPU Cluster

A self-service computing resource available to UCSC researchers who need powerful and scalable GPU-focused computing. As part of the National Research Platform (NRP), Nautilus is a distributed but centrally managed cyberinfrastructure focused on GPU-accelerated research.

Get started

Get Started with Nautilus Cluster

Features

Deploys software in containers for efficiency, resilience, scalability, and repeatability using a Kubernetes-controlled cluster
Self-managed users and user groups (namespaces)
Access to a range of GPUs, including Quadro, RTX, L40/L40s, A10, A100, and Tesla series. View current GPU resources
Supports the use of Jupyter Notebooks and Coder shared development environments, giving researchers easy access to powerful computational resources through a user-friendly interface
Self-paced training
Anyone with a UCSC CruzID Gold username and password is eligible to log on to Nautilus
Familiarity with the Unix command line and Kubernetes are required

Eligibility

Faculty, Graduate Students, Researchers, Staff, Undergraduate Students

Support

Nautilus Documentation
Nautilus Tutorials provide self-paced, basic training
Support is available through the NRP’s Matrix community chat service
Learn more about Research Computing in the ITS Knowledge Base

Cost

Nautilus is provided at no cost. Researchers who need guaranteed resource availability are encouraged to request a consultation.

Requirements

Requires an active CruzID and use of a Virtual Private Network (VPN).
Not approved for the use of Protection Level 3 (P3) data or Protection Level 4 (P4) data, including PII, HIPAA, FISMA or any other Controlled Unclassified Information (CUI)

Terms of use

Nautilus is for non-commercial research only
Users who require priority access can contribute hardware (e.g., GPU servers, FPGA, storage). Request a consultation to learn more