Special session: Managing computation, precision, accuracy and performance on exascale systems [breaker page]

ExaScale-level systems may be available in less than a decade. Computer architects are already thinking of, and planning ways to achieve such levels of performance in the near future. It can be expected that researchers and engineers will carry out scientific and engineering computations more complex than ever before, and will attempt breakthroughs not possible today. If the size of the problems solved on such machines scales accordingly, we may face new issues related to precision, accuracy, performance, and programmability. The purpose of this special session is to allow experts and others interested in ExaScale computing to present their thoughts about opportunities and challenges related to computational problems on ExaScalelevel systems. The session consists of three talks: (1)"Precision, Accuracy, Rounding, and Error Propagation" by Marius Cornea, Intel Corporation, USA; (2) "Managing the Dense Linear Algebra Software Stack" by Robert A. van de Geijn, University of Texas at Austin, USA; and (3) "Numerical Accuracy and Reproducibility at ExaScale" by Hong Diep Nguyen and Jim Demmel, UC Berkeley, USA.