The availability of high-speed networks and increasingly powerful commodity microprocessors is making the usage of clusters, or networks, of computers an appealing vehicle for cost effective parallel computing. Clusters, built using Commodity-Off-The-Shelf (COTS) hardware components as well as free, or commonly used, software, are playing a major role in redefining the concept of supercomputing. In this paper we discuss the reasons why COTS-based clusters are becoming popular environments for running supercomputing applications. We describe the current enabling technologies and present four state-of-theart cluster-based projects. Finally, we summarise our findings and draw a number of conclusions relating to the usefulness and likely future of cluster computing. Copyright © 1999 John Wiley & Sons, Ltd. KEY WORDS: commodity components; clusters; message-passing; supercomputing; parallel computing

Abstract. One-sided Communications is one of the extensions to MPI set out in the MPI-2 standard. We present here a thread-based implementation of One-sided Communications written for WMPI, an existing Windows implementation of MPI written at the Universidade de Coimbra. This is a major step towards WMPI incorporating the MPI-2 standard, with the further bene t of contributing to the thread safety of WMPI. We discuss the main design decisions associated with the implementation and consider further research work required in this area to improve both the existing implementation and to assess other implementations of One-sided Communications. 1

In this paper we evaluate the MPI environments currently available for Windows NT on the Intel IA32 and Compaq/DEC Alpha architectures. We present benchmark results for low-level communication and for the NAS Parallel Benchmarks to allow comparison to other systems, but our primary interest is determining real application performance and robustness in production cluster environments. For this we use PAFEC-FE, a large FORTRAN code for finite-element analysis. We present results from three MPI implementations, two architectures, and three networking technologies (10 Mbit/s and 100 Mbit/s Ethernet and 1 Gbit/s Myrinet).

In this paper we evaluate the MPI environments currently available for Windows NT on the Intel IA32 and Compaq/DEC Alpha architectures. We present benchmark results for low-level communication and for the NAS Parallel Benchmarks to allow comparison to other systems, but our primary interest is determining real application performance and robustness in production cluster environments. For this we use PAFEC-FE, a large FORTRAN code for finite-element analysis. We present results from three MPI implementations, two architectures, and three networking technologies (10 Mbit/s and 100 Mbit/s Ethernet and 1 Gbit/s Myrinet).* Keywords: Clusters, MPI, Windows NT 1