Preface

We are pleased to present a special issue of the best papers presented at the Scientific Day of the International Supercomputing Conference (ISC), during June 2007 in Dresden, Germany. ISC has always been an event where the academia and the industry meet to exchange ideas and shape the future of HPC. Started inMannheim, Germany, in 1986 by Prof. HansWernerMeuer, this conference has traditionally featured a diverse set of invited presentations. 2007 marked an important change in the ISC—the Scientific Day—which offered supercomputing experts for the first time an opportunity to submit papers to be presented at the event through a peer review process. A total of 45 papers were submitted for consideration and reviewed by leading experts in supercomputing and high-performance computing. A total of 16 presentations were given at the inaugural Scientific Day. Six papers were accepted for publication in this special issue of Concurrency and Computation: Practice and Experience. As is typical of the ISC itself, the paper contributions cover a wide variety of important aspects of high-performance computers. The papers from ISC07 Scientific Day presented here may be grouped into three broad categories: papers on making effective and efficient use of supercomputers; papers regarding the measurement of performance of supercomputers; and papers describing particular aspects of specialized supercomputer hardware. Three papers dealt with the topic of making effective and efficient use of supercomputers. The subject of power consumption generally is very often in the news today, and particularly so within the high-performance computing community. The paper by DiRienzo et al. describes the development of a power aware job scheduler. The work by Stewart et al. presents the experiences of the integration of a large system into a grid environment. In addition to the operational and performance issues it describes the presentation of user-friendly working environments using science gateways. The impact of migrating from a single core to a multicore CPU is analyzed by Brightwell et al. This paper presents specific features of the Cray XT3 performance measured with microbenchmarks and analyzed with respect to the observed application performance to clarify the impact of the migration of a large system to multicore CPUs. This paper was selected as an ISC award winning paper. Two papers discussed performance measurement. The paper by Mueller et al. presents an important new application benchmark suite for MPI parallel applications. This suite was developed by the SPEC High-Performance Group and is a significant contribution to the general area of supercomputer performance measurement. The features of this benchmark suite are presented in detail together with the first performance results on a number of parallel systems. The contribution of Mucci and Mohan presents the creation of a vendor-independent open source performance suite for Linux clusters. It also gives a brief overview of the current efforts in the field of tools for performance analysis for parallel programs.