As scientists continue to produce data from experiments, sensors, and simulations, the next challenge is in analyzing and visualizing the results. Addressing this issue is not just a problem of technology nor of simply deploying larger computers and data storage devices; (future) scientists need to develop the skills to effectively take advantage of the latest resources for advancing scientific discoveries.
The Remote Data Analysis and Visualization Center (RDAV) is well-situated to address essential issues in workforce development by developing and providing a robust learning opportunity for undergraduate students. RDAV students work on developing software to address computational problems stemming from active projects in science and engineering and proposed by researchers who have been allocated computation time on RDAV?s Nautilus hardware. Experience with these hardware and software resources are each, separately, uncommon skills that undergraduates seldom encounter; the combination is extremely rare. These projects integrate experience working with high performance computing hardware as well as the specialized software tools used in scientific domains. Students gain important applicable skills in computational science.
This one-year project supported in total six domestic undergraduate students (ranging from rising sophomore to rising senior) to participate in leading edge research on using High Performance Computing (HPC) for applications that require data analysis and visualization. These students came from research universities such as Princeton University and University of Tennessee, as well as a community college such as Lansing Community College in Michigan. Two are female and one is an under-represented minority. They participated in research that aimed at introducing HPC technology into fields that had not traditionally leveraged advanced computing technology but can gain immediate and significant benefits by doing so. These students worked on projects related to biodiversity, sustainability of ecosystems, and humanities and social justice. Through the project, we have confirmed that undergraduate students can effectively participate in application-driven research that uses HPC to solve specific and pressing real-world problems. As a result, students can gain much enhanced understanding of the field and move on to apply their HPC skills on projects that they had previously imagined to be beyond their capability. During each research project, these undergraduate students typically gained the following experiences and skills. - Doing research, literature review and in-person interviews with users from application fields to understand the computing needs, as well as the big picture of how the computing needs would fit into a user's overall workflow. - Learning about various visualization, data analysis and parallel computing technologies that are available on the HPC platforms operated by the RDAV center, and design an integrated prototype that serves user needs using the component technologies. - Test for correctness and performance. - Interact with users for comments and suggestions and then iterate on the design of the system. These exercises by undergraduate students, at a very low cost on both time and financial terms, allowed computer science research to penetrate into disciplines that traditionally do not use HPC but could and should use HPC technology to modernize their work. Lessons learned, although most are not at a publishable stage, have motivated more advanced current work by graduate students and research staff. Furthermore, in tangible ways, the prototypes created by these undergraduate students have already influenced how the application domain researchers would design their follow-up studies.
