Performance Evaluation of Distributed Computing over Heterogeneous Networks

RWAPI is a low-level communication interface designed for clusters of PCs. It has been developed to provide performance to higher applications on a wide variety of architectures. We implemented RWAPI on top of the modular software architecture called GRWA. RWAPI supports Ethernet, InfiniBand and Myrinet network interconnects. This paper introduces RWAPI and the design of its network component on top of both InfiniBand and Myrinet interconnects. We obtained a very low latency and high throughput compared to MPI results.

[1]  Charles L. Seitz,et al.  Myrinet: A Gigabit-per-Second Local Area Network , 1995, IEEE Micro.

[2]  Khaled Elleithy,et al.  Advances in Systems, Computing Sciences and Software Engineering: Proceedings of SCSS 2005 , 2006 .

[3]  Dave Turner,et al.  The MP_Lite Message-passing Library , 2002, IASTED PDCS.

[4]  Dan Bonachea.  GASNet Specification, v1.1 , 2002 .

[5]  Peter Thompson,et al.  IEEE P1355 - Serial Scalable Interconnect for Parallel System Construction , 1994, EUROSIM.

[6]  Alex Rapaport,et al.  Mpi-2: extensions to the message-passing interface , 1997 .

[7]  Alexandre Fenyo.  Conception et réalisation d'un noyau de communication bâti sur la primitive d'écriture distante, pour machines parallèles de type "grappe de PCs" , 2001 .

[8]  Rizos Sakellariou,et al.  Euro-Par 2001 Parallel Processing , 2001, Lecture Notes in Computer Science.

[9]  Eric Renault,et al.  A Qualitative Analysis of the Critical's Path of Communication Models for Next Performant Implementations of High-speed Interfaces , 2006 .

[10]  Alain Greiner,et al.  Protocol and performance analysis of the MPC parallel computer , 2001, Proceedings 15th International Parallel and Distributed Processing Symposium. IPDPS 2001.

[11]  Pierre David,et al.  PAPI Message Passing Library: Comparison of Performance in User and Kernel Level Messaging , 2001, Euro-Par.

[12]  Ouissem Ben Fredj,et al.  RWAPI over InfiniBand: Design and Performance , 2001, 2006 Fifth International Symposium on Parallel and Distributed Computing.