Center for Scientific Computing (CSC)
Point of contact: Paul Weakliem
Website: https://csc.cnsi.ucsb.edu
High Performance Computing (HPC) Resources:
- Free cluster options:
- Pod: most power full option providing a mixture of 40-core CPU nodes and 15 quad NVIDIA GPU
- Knot: older option, good for jobs that exceed your local desktop computer, but don't need the full performance of Pod
- Buy-in options, aka “Condo Cluster”:
- You can buy nodes from the cluster (CSC takes care of all the rest of the infrastructure.
- Default configuration: dual CPU compute node with two 24 core Intel-based CPUs with 256GB of RAM. Further options available on-demand, including GPU nodes.
- Flat cost (+ additional installation cost depending on setup request): ~ $9k
General Research IT (GRIT)
Point of contact: Michael Colee
Website: https://grit.ucsb.edu
High Performance Computing (HPC) Resources:
- Several systems available with both CPU and GPU nodes available
- Web based access available (rstudio, jupyterhub)
More details are here: https://bookstack.grit.ucsb.edu/books/hpc-usage/page/hpc-basics
Letters & Science IT (LSIT) Cloud Infrastructure
Point of contact: andreas@lsit.ucsb.edu
Website: https://cloud.lsit.ucsb.edu
Campus Community Computing Environments Resources: Infrastructure supports academic exploration (sandboxing)— instruction, research, and innovation.
Free computing options for ephemeral projects:
- Jupyterhub: vetted Jupyter environments aimed at classrooms and research labs.
- Outerrim cluster: repurposed compute providing virtual machines or containers for exploration
- Compute-list cluster: modern compute for active Proof of Concept projects
Contact LSIT for collaboration on production-level computing solutions.