Application-bypass reduction for large-scale clusters. (6th February 2006)

• Record Type: Journal Article
• Authors: Wagner, Adam; Buntinas, Darius; Brightwell, Ron; Panda, Dhabaleswar K.
• Is Part Of: International journal of high performance computing and networking. Volume 2:Number 2/3/4(2004)
• Page Start: 99
• Page End: 109
• Publication Date: 2006-02-06
• Subjects: application-bypass -- reduction -- collective communications -- process skew -- heterogeneous -- cluster computing -- MPI -- MPICH -- GM -- Myrinet; High performance computing -- Periodicals; Computer networks -- Periodicals; High performance computing; Periodicals; 004.05
• Languages: English
• ISSNs: 1740-0562

Abstract

Process skew is an important factor in the performance of parallel applications, especially in large-scale clusters. Reduction is a common collective operation which, by its nature, introduces implicit synchronisation between the processes involved in the communication and is therefore highly susceptible to performance degradation due to process skew. A collective operation with application-bypass does not require the application to block in order for the operation to make progress. Application-bypass collective operations are therefore highly tolerant of skew. In this paper, we describe the design and implementation of an application-bypass version of the reduction operation in MPICH over GM. We evaluate our implementation on a 32-node cluster. Under conditions of process skew we find a factor of improvement of up to 5.1 for our application-bypass reduction versus the default MPICH implementation. In addition, we see that this factor of improvement increases with system size, indicating that the application-bypass implementation is more scalable and skew-tolerant than the default non-application-bypass version. This framework promises design and development of high-performance and scalable collective communication libraries for next-generation large-scale clusters.