Funding: Staffing for Research Computing What staffing models does your institution use for research computing? How does your institution pay for the staffing costs associated with the spectrum of research computing activities from research/experimentation to infrastructure and sustainability? – Distributed research computing such as labs, centers,etc? – Central IT research services? What are the staffing issues you face for research computing services going forward?

The Survey findings: Staffing Summary Staffing to support research computing: Responses vary widely, mostly along these lines: – Institution/department pays – Grants pay – It depends or it’s a blend: Some grant money Some departmental money Use whatever is available Don’t worry, someone will pay

10. Staffing to support research computing: #AnswerResponse% 1NO00% 2Yes, centrally managed1680% 3Yes, departmentally managed 1680% 4 *Yes, managed otherwise (specify) 315% * Yes, managed otherwise (specify) Individual PIs Generally the researchers employ their own support staff using Post Docs and/or other IT staffing hired specifically for their research. grants Total Responses: 20

11. Are your centrally managed staff: #AnswerResponse% 1*Funded by charge back319% 2Funded centrally1594% 3 **Funded otherwise (specify) 531% **Funded otherwise (specify) grant-supported a consultancy, that funds through grants and charge back some research grant support grants Grants Total Responses: 16 *Charge Back: from Cornell SRCC breakout: “Charge for cycles, usage went from 90% to 10%

12. Are your departmentally managed staff: #AnswerResponse% 1Funded by charge back425% 2Funded departmentally1488% 3 *Funded otherwise (specify) 531% *Funded otherwise (specify) grant-supported Grant income For those staff employed directly by the researcher they are funding through the grants, contacts, and/or awards for that researcher. grants Not sure Total Responses: 16

Staffing: Funding issues Staff costs: Indirect Cost recovered or not? What’s acceptable on your campus? Are staff focused on research computing fundamentally the same/different from other IT staff? Expertise differential? Recruitment: where do you find them? Grow them? Salaries: how to determine market? Training? What new skills are needed? How to build expertise? Interface to researchers? Research IT Consultants? What do they look like? How do they work?

Staffing: Funding issues, cont’d: Staff: fixed cost or variable? Tom Sawyer model: grad & undergrad to help paint the research computing fence. Incentives to collaborate between central & distributed/labs. What are they? WIIFM? Staff administration: central or in labs? Who hires? Who manages? Succession planning?

Summary of Human Resource Recommendations: Educause/CCI Working Group Coalition for Academic Scientific Computation Developing a Coherent Cyberinfrastructure from Local Campus to National Facilities: Challenges and Strategies Workshop: July 2008 Report: February 2009 Reviewed and Discussed: March 2010

§2.4: Human Resources and Broader Impact SR 2.4.1: Agencies and campuses should support a strategic investment in human capital and curricula in order to build a pipeline of qualified experts who can develop the full capacity of CI. – TR 2.4.1a: Institutions should commit to supporting the development and delivery of modules, workshops, and courses to address the growing need for CI literacy.

§2.4: Human Resources cont’d SR 2.4.1, cont’d – TR 2.4.1b: Curricular materials for computational scientists should include systems, architecture, programming, algorithms, and numerical methods, and should prepare them to think across disciplinary boundaries. – TR 2.4.1c: National organizations and/or open- source mechanisms should be used to share curricular materials.

§2.4: Human Resources cont’d SR 2.4.2: Agencies and campuses should develop technologies and tools to use the emerging CI for education and scholarship. – TR 2.4.2a: A diverse set of communities should commit to the implementation of advanced CI technologies before there is an obvious return on investment. … – TR 2.4.2b: Investigate whether technological and organizational factors that support effective virtualization can be standardized or provided as commoditized infrastructure. … – TR 2.4.2c: Offer awards for supporting community services at all levels, including the development of new science applications, operation of technology infrastructure, and ongoing maintenance of these services. …

§2.4: Human Resources cont’d SR 2.4.3: Agencies and campuses should invest in partnerships between industry and academia. – TR 2.4.3a: These partnerships should work with businesses to adopt the use of computational science and supercomputing and assist the transfer of new computational science and supercomputing technologies from sponsored research projects to small and medium-sized businesses. – TR 2.4.3b: These partnerships should identify industry needs for new modeling software, adapt software to run effectively on modern supercomputer platforms, and provide a repository for sharing this software.

§2.4: Human Resources cont’d SR cont’d – TR 2.4.3c: Academia and industry should adopt a sensible model for sharing intellectual property. The NSF Industry/University Cooperative Research Center program could provide a viable model. – TR 2.4.3d: Academia an industry need to develop effective strategies to encourage students from traditionally underrepresented groups to pursue academic careers in computational science and to address workforce needs in industry.