Ursa Major, UCR's Google Cloud research program, groups cloud resources into tiers. The tier decides how a resource is funded. This article summarizes the tiers. The Ursa Major guidelines and, for recharged projects, your MOU, are the governing terms.
Projects set up before October 2026 may still be arranged under the earlier tiers, which covered a list of general cloud services. That arrangement is described in an archived reference article.
Tier 1: Campus-supported (no recharge to the lab under current terms)
Purpose: support research that campus systems cannot easily serve, in three areas. Tier 1 resources carry no recharge to the lab under current terms, within limits and subject to eligibility. They depend on continued campus funding and can change.
1. AI model access for research
Access to generative AI models for research and programming, through a service Research Computing manages, with a per-lab allowance. The allowance and the models offered are set when access is arranged and may change. Use beyond the allowance, and models or services outside it, are recharged.
2. Exotic hardware not available on campus
Hardware the HPCC cluster does not offer and that is hard to find elsewhere, such as Google Cloud TPUs or Arm processors, provided through a Research Computing HPC cluster in Google Cloud, within limits set per project.
Access starts with a consultation. We look at the work with you, confirm it needs hardware the HPCC does not have, and set the project's limits. Work that runs well on the HPCC belongs on the HPCC.
3. Archive storage
Google Cloud Storage archive classes (such as Coldline) for data you need to keep but rarely read. See Cloud archive and KB012: Using Ursa Major archive storage. Reading data back can carry retrieval charges.
Tier 2: Recharge
Purpose: all other cloud work.
Funding: billed to a lab funding source (COA) through ITS, at the rates in the University of California agreement with Google, under an MOU.
Tier 2 includes general-purpose cloud services, for example:
- virtual machines and research workstations;
- GKE (Kubernetes) clusters;
- Cloud SQL and other managed databases;
- BigQuery and other analytics;
- Standard-class Cloud Storage, persistent disks and Google Filestore;
- cloud GPUs and high-performance machine types used outside Tier 1;
- marketplace and third-party models through Vertex AI;
- large-scale instructional use.
See KB007: Ursa Major recharge workflow.
Tier 3: Dedicated agreement
Purpose: large or multi-year projects that need their own contract terms.
Funding: a direct agreement between the lab and Google, with ITS administrative oversight.
Regulated data: the secure research enclave
The secure research enclave is a separate environment, recharged to a grant-funded COA, for work under NIST SP 800-171 or CMMC Level 2 requirements, including controlled-access data such as NIH dbGaP. It is available after review, an approved data security plan and training.
When the cloud is not the best fit
For GPU and large batch computing, other options are often a better fit or cost the lab less. Consider them first:
- The HPCC cluster, the campus shared cluster: $1,000 per lab per year (HPCC Recharging Rates 2026/2027, as of Oct 2026).
- NRP Nautilus, for containerized GPU work on non-sensitive data.
- The NAIRR Pilot, for AI research allocations by application.
- NSF ACCESS, for national allocations by application.
Ask us if you are unsure which fits.