Research Computing FAQs
Amarel FAQs
-
Getting started is simple. First, email help@oarc.rutgers.edu, and an OARC team member will work with you to understand your research requirements and help determine the appropriate compute and storage resources for your project, including recommended specifications, quantities, and costs.
Once the details have been determined, the Business team will process an MOU (Memorandum of Understanding) outlining terms, length of ownership, costs, and other details. Upon signing the MOU and providing your account string to be charged (either GL or Project), your hardware will be allocated to you and your designated lab members. Finally, the Business team will process a journal entry transferring funds due from your account.
If you or your collaborators are not already Amarel users, you will also need to complete the General Access Account Request Form to obtain an account.
-
Amarel’s shared ownership (“condo”) model creates benefits for researchers, funding agencies, Rutgers University, and the Office of Advanced Research Computing (OARC).
Priority access to computing resources
The primary benefit of ownership is guaranteed, priority access to computing resources. While all idle resources are made available to the broader Rutgers research community through Amarel’s shared environment, owners receive preferential access to the nodes they purchase. Owners can start jobs at any time, and their jobs are not subject to preemption by other users. Their resources remain available whenever needed, except during scheduled maintenance or unexpected system outages.This model provides the experience of having dedicated computing equipment in your own lab while eliminating the challenges and costs of building and operating a private cluster. Owners do not need to manage hardware, maintain a data center, handle vendor relationships, troubleshoot system failures, or train students and staff to administer computing infrastructure. OARC manages the entire environment, allowing researchers to focus on their research.
Cost-effective use of grant funding
Amarel nodes and storage can be purchased as capital equipment, either as part of the overall cluster or in increments as small as a single node. As capital equipment, these purchases are generally not subject to Facilities & Administration (F&A) or indirect cost charges on sponsored project budgets, providing significant savings to both researchers and funding agencies.In addition, OARC purchases hardware in large volumes, securing substantial vendor discounts that are passed directly to researchers. This approach reduces costs and enables investigators from a wide range of disciplines, including those with more limited research computing budgets, to access advanced computing and storage resources.
Researchers interested in including Amarel compute or storage resources in a grant proposal should contact help@oarc.rutgers.edu for current pricing, quotes, and budgeting guidance.
Benefits to the Rutgers research community
Rutgers, like many of its peer institutions, has benefited from centralizing research computing infrastructure and expertise through OARC. The university invests in the core Amarel infrastructure, including data centers, networking, support staff, general-access compute nodes, and shared storage. Researcher investments in owned nodes help continually refresh and expand the cluster with the latest technology.Owners pay only for the specific compute and storage resources they purchase, at vendor cost, while gaining access to a professionally managed, enterprise-scale research computing environment. At the same time, the entire Rutgers research community benefits from a larger, more capable, and more sustainable shared computing ecosystem.
-
Yes, but there is the chance that a job being run by a non-owner (referred to as a General or Open Access User) may be preempted if they are on an owned node and the owner begins to run their own job.
-
Through the MOU process, while working with the Business team, owners will provide OARC with the appropriate university account strings(s), either general ledger or project. The Business team will perform a journal entry to transfer funds for your equipment when it is ready to be provisioned to you.
-
Rutgers University has purchased specialized computing, networking, and storage resources using university budgetary resources, state appropriations, endowment funds, grants, or contracts, and has designated OARC as the office responsible for managing that equipment. By gaining access to and using resources associated with OARC, you acknowledge and agree to comply with the OARC conditions of use, which can be found here.
-
Only Amarel-PE (Protected Environment), part of the Clinical and Research Data Warehouse, is suitable for working with sensitive or protected research data. The Amarel high-performance computing cluster does not meet the federal regulations necessary for compliance with NIST 800-171 or HIPAA regulations; Amarel PE does. If you will be working with protected research data of ANY type, please contact the CRDW team for a consultation.
CRDW FAQs
-
Rutgers University has made significant, strategic investments in new technologies and expertise in computational medicine and Biomedical Informatics to advance key translational and clinical projects. A cornerstone for these efforts is an enterprise-wide clinical research data warehouse and analytics platform, the Clinical and Research Data Warehouse (CRDW) that reliably integrates data arising from Electronic Medical Records (EMR); Clinical Trial Management Systems (CTMS), registries, biospecimen repositories, radiology and pathology archives and next generation sequencing devices, providing the required analytics to enable groundbreaking research.
A collaborative team from Rutgers and Rutgers Health collects and standardizes this wide-ranging clinical and research data. Leveraging expertise in data mining, computing, and pattern recognition, our goal is to tailor treatments, refine medical practices, and uncover disease insights.
This resource is invaluable for researchers, clinicians, and healthcare professionals, providing access to aggregated and organized clinical data. It facilitates research, analysis, and the generation of insights into various medical conditions, treatments, and outcomes.
-
Typically, access is granted to authorized researchers, clinicians, and stakeholders associated with Rutgers University and RWJBH, as well as their partner institutions, who have a need to work with sensitive or protected data and must abide by federal and other data protection regulations when selecting data storage and analytics resources. However, projects are subject to approval by the CRDW Data Governance Council (DGC), a board of university stakeholders and leaders who evaluate every submitted project for appropriateness.
-
The warehouse contains a wide range of clinical data, including patient demographics, medical records, laboratory results, imaging reports, and more, enabling research across various healthcare domains.
Requests for a limited data set or de-identified data (using the Safe Harbor method) can be made depending on the needs of a project and the subject of the Institutional Review Board (IRB) protocol approvals.
-
Yes, as a university core facility, CRDW is a paid service, though highly subsidized by University partners. For a list of CRDW services and their associated costs, view the current Fee Schedule, subject to change according to mandatory bi-annual core facility rate review.
-
Submit any questions regarding CRDW services by sending an email to crdw_requests@oarc.rutgers.edu.
Research Data Storage FAQs
-
RDS is intended for faculty/principal investigator-level university researchers who need a dependable, scalable storage solution for research data, including large datasets, but do not require high-performance or compute-intensive access. The platform, jointly managed by the Office of Advanced Research Computing (OARC) and OIT Enterprise Infrastructure, currently provides 34PB of usable storage capacity, with room to expand as demand grows.
Researchers who require real-time data processing or frequent data manipulation may find that Amarel, OARC’s high-performance research computing cluster, better meets their research needs. Those working with sensitive or restricted datasets, such as health-related, financial, or educational and administrative information, should consider other approved storage options.
RDS shares will not be granted to students at any level. If a lab, department, or faculty/PI researcher would like to collaborate with students within RDS, the faculty member or lab should request a share in their name and provide access to their student lab members.
-
RDS provides reliable, highly available, high-capacity storage for research data of any size. It offers a simple and consistent way for researchers to store, access, and share their data, ensuring that research data remains secure and accessible over time.
-
Purchase agreement terms of use
- To ensure responsible data management, retention, and eventual removal, all allocations will be managed by a Principal Data Share Holder, and all shares will require agreement of oversight by a Dean, Director, or Department Head in the case of data left behind when a Share Holder departs the university.
- Collaborative or departmental shares are also available when desired—the designated Principal Share Holder will assign individual user permissions in this case and is responsible for compliance with all sponsored project and other data regulations.
- Costs and utilization will be reassessed by OIT annually and adjusted if necessary.
- Details of the purchase agreement will be provided in a Memorandum of Understanding (MOU), which outlines the roles, responsibilities, and mutual expectations between the Office of Information Technology (OIT) and the Principal Data Share Holder.
-
RDS offers secure, scalable, and cost-effective storage for non-sensitive research data that exceeds the capacity of conventional storage solutions. Researchers receive 2TB of storage at no cost, with the option to purchase additional space in 1TB increments for $61.44 per TB per year.
Highlights of billing terms
- Data storage beyond 2TB can be requested in TB increments and will be charged annually at $61.44 per TB per year ($0.06/GB/year).
- Annual billing will take place in January and first year allocations will be prorated accordingly, depending upon when the initial share is requested.
- Billing is determined by the number of TBs requested for the year, not by actual usage.
- ShareHolders may purchase multiple years in advance if in compliance with the terms of their sponsored projects.
-
To get started, fill out the Research Data Storage Access Request Form. Once received, a team representative will be in touch to work with you.
-
Box is a cloud storage, meant for general files, documents, pictures, etc. RDS is an onsite storage platform intended for research data. It can handle large files and many small files, etc. It’s not meant for office documents or high-performance computing (HPC) workloads.
In addition:
- Use RDS if want your data on-prem and use Box to store your data in the cloud.
- Use Box for protected health information (PHI). (PHI can be stored in Box provided it is stored in a restricted folder).
- Use RDS if you need to store more than 5TB.
To further compare the two platforms, select them on the Data Storage Finder matrix found at researchdatafinder.rutgers.edu and scroll down. (Users must be connected to the Rutgers network or VPN to access).
-
For additional information on the RDS platform, please email research_storage@oarc.rutgers.edu.
Sponsored project collaboration FAQs
-
Contact the OARC Business Office for all Amarel quotes, including those for grant submission proposals.
-
Absolutely. OARC is here to assist the university research community with long-term strategic research computing, network, storage, and data-science infrastructure planning, including those enabled by sponsored programs. We have successfully brought in such infrastructure with NSF, NIH, and other funded research dollars for the benefit of the entire university, and we will continue to support such efforts in cooperation with our Enterprise Infrastructure partners as part of our mission.
-
No, you do not need to become an Amarel owner to include OARC resources in your grant submission. Many users are general access, whose research benefits greatly from utilizing the Amarel condo model at no cost, and/or work with our research scientists to improve research output.
-
OARC has created a file with a summary of our compute, storage, and networking resources which you can submit as your Facilities and Resources document in your grant proposal submission. We ask that you let us know at the time of proposal if you will be purchasing or utilizing OARC equipment using award funds, should your proposal be funded, so we can plan ahead with Enterprise Infrastructure. If you need a quote for your proposal budget, see info here.
-
OARC’s Senior Research Scientists are made up of primarily PhD-level domain scientists who are well-versed in writing proposals, performing research, and analyzing data in collaboration with PIs and their teams. Often, our REFs are written into grant budgets to account for their contribution to funded projects. They have expertise in life sciences, GIS/visualization, STEM disciplines, clinical and medical informatics, digital humanities, and social sciences. Please work with the OARC Research Partnerships team if you are considering including OARC scientists and staff members in your proposal. In addition, we invite you to contact the OARC Senior Research Scientists team at help@oarc.rutgers.edu for individual or group training, consultation, and assistance in putting Amarel to work for effective and cutting-edge research.
-
Please reference OARC or the Amarel cluster in any research report, journal, or publication that requires citation of an author’s work. Recognizing the OARC resources you used to conduct your research is important for our process of acquiring funding for hardware, support services, and other infrastructure improvements. The minimal content of a reference should include:
Office of Advanced Research Computing (OARC) at Rutgers, The State University of New JerseyA suggested acknowledgement is:
The authors acknowledge the Office of Advanced Research Computing (OARC) at Rutgers, The State University of New Jersey for providing access to the Amarel cluster and associated research computing resources that have contributed to the results reported here. URL: https://it.rutgers.edu/research-computing
