SESSION LEADER(S):Martin Herbordt, Alan George, Herman Lam
ABSTRACT: Reconfigurable computing is characterized by hardware that adapts to match the unique needs of each application, which offers unique advantages in speed per unit energy. With a proven capability of 1 PetaOPS at 10K Watts, large-scale reconfigurable supercomputing has an important role to play in the future of high-end computing. One such system, the Novo-G machine in the NSF CHREC Center at the University of Florida, has a user base of over a dozen research groups distributed worldwide and is rapidly moving towards production utilization in Bioinformatics and other scientific and engineering domains. The purpose of this BoF is to introduce the basic architecture of such systems, describe applications and tools being developed for them, and provide a forum for discussing emerging opportunities and issues in terms of performance (speed and scalability), productivity (programmability and portability), and sustainability (energy and TCO).
Session Leader Details:
Martin Herbordt (Primary Session Leader) - Boston University
Alan George (Secondary Session Leader) - University of Florida
Herman Lam (Secondary Session Leader) - University of Florida