High Performance Message Passing Library for Maestro2 Cluster Network

K. Aoki, K. Wada, M. Ono (Japan), and S. Yamagiwa (Portugal)


cluster computing, high-performance network, message passing library, performance evaluation


A cluster has become a popular platform for high per formance computing. Currently, many PC clusters are organized with cost-effective conventional wide-area network technologies to connect each processing elements. Here, it is difficult to extract potential performance of the cluster, since there is the disparity between the wide-area network technologies and the performance characteristics required by cluster computing. We have developed a net work to build high performance PC cluster, called Maes tro2 cluster network. This paper describes about high per formance communication library for Maestro2, named MMP. First, we describe the novel technique, design, and implementation of Maestro2. Then, we present a design and an implementation of MMP. We also show experi mental results to evaluate the basic performance of MMP and compare it with raw performance of Maestro2. The experimental results show that MMP can extract maxi mum performance from Maestro2.

Important Links:

Go Back