What are we trying to do?
We're trying to "energize" high performance computing research and education. Below is the project summary from a coming NSF proposal. The "proposed activity" is to expand the lab to reach out to more SVSU faculty, undergraduates and the general public.
PROJECT SUMMARY
Overview – The proposed activity intends to advance research and enhance undergraduate education at Saginaw Valley State University (SVSU) by expanding faculty research and classroom lessons in algorithm/simulation development to include theoretical concepts in parallel processing and practical skills in high performance computing (HPC). The project aims to build on existing hardware: (1) to discover new application areas and research domains to apply parallel programming and Field Programmable Gate Arrays (FPGAs), (2) to energize student engagement with collaborative research and teaching, (3) to develop student and faculty hands-on skills using shared HPC platforms, and (4) to inspire and expand an FPGA research community where faculty, undergraduates, industry and the broader community can participate and collaborate.
SVSU already has an FPGA lab. Each faculty participant already leads existing research using non-FPGA methods. The proposed project would integrate existing faculty research to include FPGAs. Faculty would collaborate, regularly and formally, in research publication, classroom lesson planning, continuing process improvement, and outreach events. Existing equipment, including multi-core CPUs, GPUs and FPGAs, would provide a shared platform for researching and teaching parallelization on high performance systems. Some research would be organized into classroom lessons and fun demonstrations accessible to lay people which would be presented at venues that reach underrepresented populations.
Keywords: FPGA; high performance computing; parallelism; algorithm development
Intellectual Merit – FPGAs have increased in computational power faster than conventional CPUs. As their power, flexibility and ease of use increases, FPGAs become more attractive for new applications, including both industrial and research applications. The proposed activity would produce more FPGA-related research papers, develop new classroom lessons, and energize the research environment at SVSU. Parallelization techniques have not been thoroughly explored in the scholarly literature of the research areas under study in this proposal. Augmenting current research to include parallel processing techniques might add new depth to the respective fields of inquiry. At the moment it is infeasible to add a new required course in parallel processing and high performance computing the CS/EE curriculum at SVSU. Integrating parallelism as part of multiple courses enables CS/EE to introduce this fundamental computing concept to its undergraduates. Above all, by working together, professors can improve research and pedagogical productivity. As time progresses, professors would be expected to exceed the average number of scholarly papers published, to exceed the average impact factor of publications compared to prior years, and to increase the number of papers published by two or more faculty members. This would be accomplished by formal monthly research reviews and frequent informal knowledge sharing.
Broader Impacts – During this project the number of undergraduate co-authors of research papers would be expected to increase, which could lead to an increase in undergraduates applying for graduate school. Advanced conceptual understandings and hands-on programming skills could encourage undergraduates to apply for professional positions in industry. A reputation for CS/EE success, coupled with outreach to the community, could inspire future researchers from underrepresented backgrounds. Lessons developed as a result of this work would be made available online, for use by professors and students at universities worldwide, or even by the general public. The faculty collaboration techniques used in this work could yield value beyond the CS/EE departments at SVSU. They might apply to other areas of collaborative scholarly inquiry, not just CS/EE, and might themselves be subjects of investigation for researchers interested in models of personnel organization in research and teaching.
PROJECT SUMMARY
Overview – The proposed activity intends to advance research and enhance undergraduate education at Saginaw Valley State University (SVSU) by expanding faculty research and classroom lessons in algorithm/simulation development to include theoretical concepts in parallel processing and practical skills in high performance computing (HPC). The project aims to build on existing hardware: (1) to discover new application areas and research domains to apply parallel programming and Field Programmable Gate Arrays (FPGAs), (2) to energize student engagement with collaborative research and teaching, (3) to develop student and faculty hands-on skills using shared HPC platforms, and (4) to inspire and expand an FPGA research community where faculty, undergraduates, industry and the broader community can participate and collaborate.
SVSU already has an FPGA lab. Each faculty participant already leads existing research using non-FPGA methods. The proposed project would integrate existing faculty research to include FPGAs. Faculty would collaborate, regularly and formally, in research publication, classroom lesson planning, continuing process improvement, and outreach events. Existing equipment, including multi-core CPUs, GPUs and FPGAs, would provide a shared platform for researching and teaching parallelization on high performance systems. Some research would be organized into classroom lessons and fun demonstrations accessible to lay people which would be presented at venues that reach underrepresented populations.
Keywords: FPGA; high performance computing; parallelism; algorithm development
Intellectual Merit – FPGAs have increased in computational power faster than conventional CPUs. As their power, flexibility and ease of use increases, FPGAs become more attractive for new applications, including both industrial and research applications. The proposed activity would produce more FPGA-related research papers, develop new classroom lessons, and energize the research environment at SVSU. Parallelization techniques have not been thoroughly explored in the scholarly literature of the research areas under study in this proposal. Augmenting current research to include parallel processing techniques might add new depth to the respective fields of inquiry. At the moment it is infeasible to add a new required course in parallel processing and high performance computing the CS/EE curriculum at SVSU. Integrating parallelism as part of multiple courses enables CS/EE to introduce this fundamental computing concept to its undergraduates. Above all, by working together, professors can improve research and pedagogical productivity. As time progresses, professors would be expected to exceed the average number of scholarly papers published, to exceed the average impact factor of publications compared to prior years, and to increase the number of papers published by two or more faculty members. This would be accomplished by formal monthly research reviews and frequent informal knowledge sharing.
Broader Impacts – During this project the number of undergraduate co-authors of research papers would be expected to increase, which could lead to an increase in undergraduates applying for graduate school. Advanced conceptual understandings and hands-on programming skills could encourage undergraduates to apply for professional positions in industry. A reputation for CS/EE success, coupled with outreach to the community, could inspire future researchers from underrepresented backgrounds. Lessons developed as a result of this work would be made available online, for use by professors and students at universities worldwide, or even by the general public. The faculty collaboration techniques used in this work could yield value beyond the CS/EE departments at SVSU. They might apply to other areas of collaborative scholarly inquiry, not just CS/EE, and might themselves be subjects of investigation for researchers interested in models of personnel organization in research and teaching.