Input Output Intensive Massively Parallel Computing

Massively parallel processing is currently the most promising answer to the quest for increased computer performance. This has resulted in the development of new programming languages and programming environments and has stimulated the design and production of massively parallel supercomputers. The efficiency of concurrent computation and input/output essentially depends on the proper utilization of specific architectural features of the underlying hardware. This book focuses on development of runtime systems supporting execution of parallel code and on supercompilers automatically parallelizing code written in a sequential language. Fortran has been chosen for the presentation of the material because of its dominant role in high-performance programming for scientific and engineering applications.

The major research results from the Scalable Input/Output Initiative, exploring software and algorithmic solutions to the I/O imbalance. As we enter the "decade of data," the disparity between the vast amount of data storage capacity (measurable in terabytes and petabytes) and the bandwidth available for accessing it has created an input/output bottleneck that is proving to be a major constraint on the effective use of scientific data for research. Scalable Input/Output is a summary of the major research results of the Scalable I/O Initiative, launched by Paul Messina, then Director of the Center for Advanced Computing Research at the California Institute of Technology, to explore software and algorithmic solutions to the I/O imbalance. The contributors explore techniques for I/O optimization, including: I/O characterization to understand application and system I/O patterns; system checkpointing strategies; collective I/O and parallel database support for scientific applications; parallel I/O libraries and strategies for file striping, prefetching, and write behind; compilation strategies for out-of-core data access; scheduling and shared virtual memory alternatives; network support for low-latency data transfer; and parallel I/O application programming interfaces.

This book contains four review articles in the area of scalable computing. Two of the articles discuss methods and tools for the parallel solution of irregular problems, which have been satisfactorily worked out in heterogeneous systems. One surveys the technology and applications of multimedia server clusters, which are playing an increasing role in the current networked environment. An additional article discusses SilkRoad, which adds distributed shared memory capabilities to the Cilk parallel programming system. Once again, the book represents a new set of steps forward in parallel systems.

This book constitutes the refereed proceedings of the Fourth International Conference on Parallel Computing Technologies, PaCT-97, held in Yaroslavl, Russia, in September 1997. The volume presents a total of 54 contributions: 21 full papers, 20 short papers, 10 posters, and three tutorials. All papers were selected for inclusion in the proceedings from numerous submissions on the basis of three independent reviews. The volume covers all current topics in parallel processing; it is divided into sections on theory, software, hardware and architecture, applications, posters, and tutorials.

This book constitutes the refereed proceedings of the Third International Euro-Par Conference, held in Passau, Germany, in August 1997. The 178 revised papers presented were selected from more than 300 submissions on the basis of 1101 reviews. The papers are organized in accordance with the conference workshop structure in tracks on support tools and environments, routing and communication, automatic parallelization, parallel and distributed algorithms, programming languages, programming models and methods, numerical algorithms, parallel architectures, HPC applications, scheduling and load balancing, performance evaluation, instruction-level parallelism, database systems, symbolic computation, real-time systems, and an ESPRIT workshop.

The book is aimed at graduate students, researchers, engineers and physicists involved in fluid computations. An up-to-date account is given of the present state of the art of numerical methods employed in computational fluid dynamics. The underlying numerical principles are treated with a fair amount of detail, using elementary methods. Attention is given to the difficulties arising from geometric complexity of the flow domain. Uniform accuracy for singular perturbation problems is studied, pointing the way to accurate computation of flows at high Reynolds number. Unified methods for compressible and incompressible flows are discussed. A treatment of the shallow-water equations is included. A basic introduction is given to efficient iterative solution methods. Many pointers are given to the current literature, facilitating further study.

This book constitutes the strictly refereed proceedings of the 9th International Conference on Computer Aided Verification, CAV '97, held in Haifa, Israel, in June 1997. The volume presents 34 revised full papers selected from a total of 84 submissions. Also included are 7 invited contributions as well as 12 tool descriptions. The volume is dedicated to the theory and practice of computer aided formal methods for software and hardware verification, with an emphasis on verification tools and algorithms and the techniques needed for their implementation. The book is a unique record documenting the recent progress in the area.