Special Issue on Advances in High Performance Computing and Simulation
Waleed W. Smari, Sandro Fiore, Mads Nygaard · Concurrency and Computation Practice and Experience · 2012
This special issue contains seven papers representing recent advances in the area of high performance computing and simulation.These papers were carefully selected from the 2009 International Conference on High Performance Computing and Simulation, which was held in Leipzig, Germany, June 21-24, 2009.The invited papers in this special issue represent augmented works originally presented at the conference which cover some of the main contemporary topics and challenges.The papers presented in this special issue can be classified in three categories: surveys, performance, and high-performance computing (HPC) for eScience.The first category includes a survey on hardware-aware and heterogeneous computing on multicore processors and accelerators [1] and a contribution on trends and challenges in operating systems [2].The performance category proposes challenging topics such as reliability bottlenecks in integrated parallel fault-tolerant systems [3], input/output in web servers [4], and model checking and code generation for transaction processing software [5].Finally, HPC for eScience presents an application to life sciences simulations based on parallel stochastic simulations with rigorous distribution of pseudo-random numbers [6] and a fast seismic modeling and reverse time migration on a graphics processing unit (GPU) cluster [7].A short description introducing all of the papers of this special issue is presented in the following paragraphs.Buchty et al.[1] present a survey on current multicore and accelerator technologies.The authors outline architectural features and show how these features are exposed to the programmer and how they can be beneficially utilized in the application-mapping process.Moreover, the necessity of hardware-aware computing and the challenges arising from high-performance heterogeneous computing are motivated and summarized.Also investigated is the interaction between hardware and application characteristics for selected applications in numerical simulations.Polze [2] presents major trends and challenges in operating systems, moving from parallel to cloud computing.In his work, the author highlights and discusses roots, design options, and stateof-the-art solutions targeting challenging topics such as virtualization, multicore systems, energy awareness, and cloud computing.In his paper, Fechner [3] tries to answer several questions concerning fault propagation in multithreaded, multicore, and manycore systems.Typically, such integrated systems have a common, shared interface to the outside world, bearing the potential of a single point of failure.A threestage model is also presented and discussed by the author to help identify weaknesses of integrated parallel fault-tolerant systems.Azzedin and Al-Issa [4] present the design of a new, self-adapting Web server architecture that makes decisions on how future input/output operations would be handled based on load conditions.The results obtained from the authors implementation indicate that it is capable of providing competitive performance and better utilization than comparable nonadaptive Web servers at different load levels.In [5], Mentis and Katsaros present a model-driven approach for generating a provably correct implementation of the transaction model of interest.The model is specified by state machines that represent the transaction participants who are synchronized on a set of events.Moreover, the authors checked various possible execution paths of the synchronized state machines for property violations and verified the specification of nested transactions to demonstrate the validity of the adopted approach.