This institutional infrastructure (II) project, called INSpiRE, is aimed at building a new shared research facility for conducting multidisciplinary research in several areas of computer science and engineering. The proposed infrastructure is a heterogeneous cluster computing system with a heterogeneous memory and storage hierarchy. This equipment will facilitate implementation, evaluation and validation of research projects that focus on several layers of the system design stack, with the common theme of leveraging heterogeneity in each of these layers for designing heterogeneous multicore architectures. It would enable cross-pollination of ideas and tighter integration across these projects for initiating new cross-cutting research. In addition, existing projects in three application domains (neuromorphic vision processing, high performance soft computing, and statistical analysis of privacy-preserving datasets) will directly benefit from this infrastructure.
The proposed infrastructure would support the research activities of more than 50 Ph.D. students in three major disciplines: hardware architecture, system software, and applications. The infrastructure will be a catalyst to advance the state-of-the-art in designing future multicore systems. Undergraduate, graduate, minority and women students involved in the project will get hands-on training in several cross-cutting areas. The tools techniques developed in this research will be used in teaching existing and new courses, and will be made publicly available.
