Skip to content

Overview

ScienceCluster is a managed high-performance computing (HPC) environment designed to support a wide range of research workloads. Whether your tasks are CPU-intensive, require large shared memory, benefit from GPU acceleration, or depend on high-speed interconnects, ScienceCluster enables you to choose the optimal resources for your needs. It consists of a network of interconnected machines that function as a single, unified system.

ScienceApps is a browser-based interface to ScienceCluster that allows researchers to run interactive applications—such as Jupyter, RStudio, and VSCode—directly from their web browser. It simplifies access to HPC resources for exploratory data analysis, visualization, and code development.

Accessing the service

Hint

Learn more about how to get access to the ScienceCluster.

Once you have a ScienceCluster user account, you can connect from the command line with your UZH shortname and Active Directory (AD) password:

ssh -l shortname cluster.s3it.uzh.ch

After running the command, you will be prompted for your password. Please note that there will be no echo as you type your password -- nothing will be displayed when you type each character.

Attention

For security reasons, you can access ScienceCluster only from the UZH internal network. Please use the UZH VPN if you are connecting from off-campus.

To enable passwordless authentication with ssh key pairs, please see here.

Best Practices Checklist

To optimize your ScienceCluster workflows and prevent data loss, consider the following best practices:

  • Use Git to version control and secure a remote copy of your code. If you don't know Git, consider ZI's trainings on it and related Science IT offerings.

  • Create an geographically redundant copy of your raw data (and various checkpoints of your outputted data) on a non Science IT system (to follow a 3-2-1 back up rule). For more information on archive services at UZH, see the corresponding FAQ page (UZH login required).

  • Take the Science IT trainings to best ensure your workflow optimizes within Science IT infrastructure. Likewise operate with deliberation on the Command Line (i.e., use rm commands with proper precaution, don't universally open read, write, and/or execute permissions for other users across all of your files, etc.).

Please remember: as a research computing cluster offered via an Infrastructure-as-a-Service model, you as a user ultimately maintain responsibility for the preservation of your research data and code. If you have any questions, you can always contact us.