Connecting
Table of Contents
Overview
This page serves as a comprehensive guide for users to understand how to access the chip cluster, what credentials are required, and what to expect in their initial environment. The chip cluster is the sole High Performance Computing Facility (HPCF) at UMBC as of April 2025.
Requesting a User Account on chip
To access the chip cluster, every user account must be part of a group affiliated with a lab, course, or special project. These cluster groups can only be requested by faculty or staff.
How to Request an Account
If you have not logged in to chip before, you must request an account by creating a ticket (Research Computing Action: High Performance Cluster, HPC Action: User Account Creation/Modification). Detailed instructions can be found here:
How to Request a Chip Account.
Information Needed for Request
For New Groups (Faculty/Staff)
You will need to indicate if the group is for a class or a research lab. For classes, provide the class name; for research, provide a project title and abstract. The requester is automatically provisioned as a user within the new group.
To Join an Existing Group (Students)
Select "High Performance Computing" under "Research Computing Action" and "User Account Creation/Modification" under "HPC Action". You will need to provide your PI's UMBC email (the group owner), the project title, project abstract, and relevant fields of research.
Approval Process
Your request ticket is sent to DoIT staff and awaits approval from the group owner. Once approved, your chip account should be created within three business days.
Note: If you previously had access to the taki or ada clusters, you likely already have an account on chip.
Accessing the chip Cluster
Click here for information on how to access Chip from your computer: Accessing Chip
To check your home directory usage: How to check home directory usage
HPCF Research Volumes (/umbc/rs/$GROUP)
In addition to your home directory, chip provides research volumes with significant storage space, typically starting at 250GB or 1TB for new groups.
All users within a PI's group have access to these volumes.
These volumes are where you should store the majority of your data and projects.
Structure
For more information on how the User Environment is set up: User Environment
Concurrent SSH Connections
Users can establish concurrent SSH connections to chip from multiple terminals or command prompts.
Login Node Resource Limit
The cumulative hardware resources (core usage, memory) consumed by your SSH sessions on the login node must remain below 2 GiB.
VSCode Server Connections
Running multiple instances of VSCode server connections can cause issues, and system administrators may kill these instances as needed.
Related Resources
For more in-depth guidance on using the chip cluster, please refer to the following helpful pages and tutorials:
Quick Run Guide: How to run things on chip, with/without parallelism, with different compilers, etc.
How to Edit Code on chip (including Emacs, Nano, Vim, VSCode)
Available Software and Packages on chip (Not University-wide)
Storage (including home directories, research volumes, and scratch space)
Back to Main Page