MRI: Acquisition of the GLOW Distributed Computing System at the University of Wisconsin - Madison

University Of Wisconsin-Madison, Madison WI

Investigators

Abstract

This project, harnessing the power of commodity hardware and software resources in a Grid laboratory (GLOW), aims at making Grid computing an effective tool for scientific research in a large-scale environment with real-life users and applications. While training a new generation of interdisciplinary scientists, this multidisciplinary team collaborates in the development, implementation, test, and deployment of Grid enabled capabilities. GLOW spans six domains: Chemical Engineering, Chemistry, Computer Science, Medical Physics (including Radiology and Human Oncology), Particle Physics and Astrophysics (each with significant computational needs and each contributing a site to the lab). The environment permits exploration on harnessing new computing technologies to meet the computational needs of leading edge research in the biological and physical sciences. Research involves: Optical mapping Neutrino observatory: IceCube Software development: Higgs boson Overcoming bottlenecks in molecular study of complex fluid and materials Applications in intensity modulated radiation therapy, IMRT Sociological nature of ownership The first technique has emerged as a powerful system for the construction of high-resolution restriction maps from a broad range of clone types and genomic DNA. The second forms part of an international initiative utilizing the South Pole ice instrumented at depths with optical sensors as a Cherenkov detector; the telescope collects the light from secondary particles produced in interactions of high-energy neutrinos near the instrumented volume. The third explores fundamental physics at a new energy regime predicted to lead to the discovery of Higgs boson that is responsible for providing masses to elementary particles or new symmetries, matter forms, or interactions. Software under development will trigger system simulation, event filter algorithms, and simulated data analysis. In four, the Chemical Engineering group conducts computational research on designing molecules and materials for specific application, focusing on two areas: development of accurate, transferable force fields capable of describing interactions between distinct atomic species in condensed media and advanced simulation techniques for describing the structure and dynamics of complex fluids and materials. Fifth, the Medical Physics, Radiology, and Human Oncology groups focus on various areas; applications ranging from Monte Carlo calculations of radiation fields produced by radiotherapy treatment units to dose calculation within a patients. Image science uses complex, high performance, iterative, back-projection techniques to develop 3-dimensional time varying images for advanced diagnosis that combined with stochastic optimization provides a new framework for clinical analysis and treatment. Lastly, social analysts will study new ownership and cooperation relations as these emerge in the areas of scientific exploration. GLOW enables the cross-fertilization of active research pursuits providing a real-life Grid environment for experimental work opening many training opportunities for students and postdocs.