Fast Collective Communication Algorithms for Reflective Memory Network Clusters
نویسندگان
چکیده
1 This research is supported in part by an NSF Career Award MPI-9502294, NSF Grant CCR-9704512, an Ameritech Faculty Fellowship award, and grants from the Ohio Board of Regents. Abstract: In this paper, we present algorithms for efficiently implementing three collective communication operations on reflective memory network clusters: Broadcast, Barrier Synchronization and All-Reduce. These algorithms have been implemented as an extension to the Bill-Board Protocol, a low-latency communication protocol for reflective memory networks that we previously developed [6,7]. Simulated performance results are presented for these algorithms. The performance of the algorithms is evaluated in detail. A comparative study with Myrinet, a popular point-to-point switched interconnect for clusters is also presented.
منابع مشابه
Efficient Barrier Using Remote Memory Operations on VIA-Based Clusters
Most high performance scientific applications require efficient support for collective communication. Point-to-point message-passing communication in current generation clusters are based on Send/Recv communication model. Collective communication operations built on top of such point-to-point message-passing operations might achieve suboptimal performance. VIA and the emerging InfiniBand archit...
متن کاملFast Collective Operations Using Shared and Remote Memory Access Protocols on Clusters
This paper describes a novel methodology for implementing a common set of collective communication operations on clusters based on symmetric multiprocessor (SMP) nodes. Called Shared-Remote-Memory collectives, or SRM, our approach replaces the point-to-point message passing, traditionally used in implementation of collective message-passing operations, with a combination of shared and remote me...
متن کاملScalable and High Performance Collective Communication for next Generation Multicore Infiniband Clusters
High Performance Computing is enabling rapid innovations spanning several key areas ranging from science, technology and manufacturing disciplines to entertainment and financial markets. One computing paradigm contributing significantly to the outreach of such capabilities is Cluster Computing. Cluster computing involves the use of multiple Commodity PCs interconnected by a network to provide t...
متن کاملAn Evaluation of the DEC Memory Channel Case Studies in Reflective Memory and Cooperative Scheduling
With the rise of clusters as a vehicle for very-highperformance computing, an increasing emphasis is being placed upon the communication interface between each processor and the underlying network. And while many studies have explored the design of both traditional send/receive network interfaces and shared-memory designs, a third alternative — reflective memory — offers a significantly differe...
متن کاملEfficient Collective Operations Using Remote Memory Operations on VIA-Based Clusters
High performance scientific applications require efficient and fast collective communication operations. Most collective communication operations have been built on top of point-to-point send/receive primitives. Modern user-level protocols such as VIA and the emerging InfiniBand architecture support remote DMA operations. These operations not only allow data to be moved between the nodes with l...
متن کامل