Advances in high performance computing

Abstract

This special issue brings together five research papers providing a snapshot of the current state of the art in the broad area of high performance computing. These papers have been selected as extended and revised versions, subject to peer review, from those originally accepted for presentation at the 3rd International Symposium on Parallel and Distributed Processing and Applications (ISPA’2005), which was held during 2–5 November, 2005 in Nanjing, China. The selected papers address a range of issues in high performance computing and their contributions are summarised below. In “An Energy-Efficient Protocol for Data Gathering and Aggregation in Wireless Sensor Networks,” Liu, Cao, Zheng and Xie introduce a hierarchical clustering protocol for data gathering and aggregation in wireless sensor networks. Their simulation results show that their protocol outperforms two existing ones, LEACH and PEGASIS, in terms of network lifetime and the amount of data gathered. More and more of the top 500 super-computers are cluster-based. However, the increased reliance on super-computing clusters based on commercial off-the-shelf parts has created a growing need for lightweight fault tolerance. In their article, “ER-TCP: An Efficient TCP Fault-tolerance Scheme for Cluster Computing,” Shao, Jin, Cheng and Jiang present a scheme that combines active replication with a logging mechanism to achieve fault tolerance for the TCP connections on the server-side. In “Improving the Parallelism of Iterative Methods by Aggressive Loop Fusion,” Xue, Guo and Dai present a compiler approach to aggressively fusing loop nests