Campus AI tools
ITS maintains the campus guidance on generative AI tools available to UCR faculty, staff and students, including which tools are licensed, how to sign in, training materials, and guidance on appropriate use. Go to AI at UCR (ITS) for the current list and for the terms that apply to each tool.
Research Computing does not restate those terms here. Before using any AI tool with research data, check the ITS guidance and your data's protection level. Your data security plan or agreements may restrict it further.
AI research at UCR
The RAISE Institute (Riverside Artificial Intelligence Research and Education) brings together UCR researchers working on AI and its applications, and runs seminars and workshops.
AI computing
| You want to | Where |
|---|---|
| Train or fine-tune models on GPUs | HPCC cluster, or national allocations (NAIRR Pilot, NSF ACCESS) |
| Call generative AI models from your research code | Ursa Major Tier 1 AI model access, with a per-lab allowance |
| Use Google's AI platform services (such as Vertex AI) directly | A recharged Ursa Major project |
| Run open-source language models yourself | Running local LLMs with Ollama, on the HPCC or a workstation |
| Use containers and notebooks with GPUs | NRP Nautilus |
Cloud for research
UCR researchers reach the major cloud providers in two ways:
- Ursa Major, UCR's Google Cloud research program, with campus-supported AI model access, exotic hardware and archive storage, and recharged projects for other cloud work.
- Cloud accounts under University of California agreements with AWS, Google Cloud and Azure, billed to your funds through ITS.
Cloud can be the right answer when you need managed services, elastic scale for a short time, or tools that do not exist on campus. For long-running batch or GPU work, the HPCC usually costs the lab less. Ask us to compare for your case.