Institute for Computational Research
in Engineering and Science
ICRES

• to enable interdisciplinary computational science and engineering research
• to promote the computational research paradigm throughout KSU
• to serve as a catalyst for the development of interdisciplinary computational research graduate programs

• provide an environment for collaborative computationally-based research
• develop computational research methods
• develop computational research software and algorithms
• develop computational research computer systems
• enable researchers campus-wide to utilize these resources without being an expert in computing systems

• secure extramural funds to promote the computational research paradigm through leadership of interdisciplinary research teams
• develop a cluster computer which be dynamically configurable to a research problem
• interconnect this cluster computer to the National Computational Science Alliance Grid
• allow users to "buy-in" to this cluster computer with contribution of equipment (for short or long periods of "buy-in")
• provide software environment which is common to national labs and supercomputer centers (but probably much less power: good for development and small experiments, but we should send the "big" jobs off to the Alliance Grid)
• provide software frameworks which enable scientists and engineers to build the solutions to their research problems without becoming an expert in distributed or parallel computing, visualization, data mining, database management, etc.
• develop BS, MSE and MS internships in CS to sponsor computational science and engineering projects for support of engineers and scientists throughout the University

• Build Beowulf computing cluster from selected groups across KSU
• CIS has developed a 24 processor (four quads and four duals) system interconnected with gigabit ethernet
• CIS will add 12 more processors to the cluster in Summer 2000
• CNS will add 12 more to the network in Summer 2000
• Wal-Mart/NCR will add a large parallel processing capability (80 processors) and large storage subsystem (1.6 terabytes)
• Once this is installed, interconnect to the Beowulf cluster via gigabit ethernet
• Once incorporated into the database and data mining research program and curriculum, acquire the software to incorporate the 80 processors into the Beowulf cluster
• solicit researchers on campus who have large workstations to include their computational power in the Beowulf cluster

• incorporate New Real-time Embedded Systems Laboratory
• extend the new Knowledge Discovery and Database Laboratory, to include visualization and virtual reality facilities
• extend Mobile Robotics Laboratory to include other disciplines
• develop Software Engineering Learning Laboratory

• BS, MSE, and MS students to enable the development of quality software in support of a variety of disciplines, especially in Engineering and Science
• develop software frameworks to enable scientists and engineers to build software solutions without becoming experts in software architecture for parallel and distributed systems, data mining, database management, real-time systems, software verification, object-oriented modeling and development, etc.
• large simulations for science and engineering
• real-time control software
• database and data mining frameworks
• parallel algorithms and systems (simulations, biological sequence analysis, functional genomics, optimization problems, computational chemistry, computational physics, etc.)
• software engineering environments and CASE tools
• parallel domain-specific problem solving environments (PSEs) (build the analog of spreadsheets for specific domains, like grid simulations, n-body problems, systems of equations solvers, finite element codes, computational chemistry, materials science simulations, etc.)