Understanding the performance of DSM applications

Asynchronous communication using fast, low latency networks present the challenge of delivering the performance such networks exhibit, from the network layer up to user level applications. Such applications which need to retrieve messages from the network, using polling based communication packages, often face the problems of responsiveness, cpu utilization and redundant polling. Because of the inherent asynchronous communication nature of software dsms, these overheads become apparent when the dsm is implemented on top of such communication packages.

Proceedings of the 7th workshop on Workshop on languages, compilers, and run-time support for scalable systems - LCR '04, 2004

Scientific applications contain program sections that exhibit repetitive data accesses. This paper proposes combined compiletime/runtime data reference analysis techniques that exploit repetitive data access behavior in both regular and irregular program sections. We present a new compiler algorithm to detect such repetitive data references and an API to an underlying software distributed shared memory (software DSM) system to orchestrate the learning and pro-active reuse of communication patterns. We evaluate the combined compile-time/runtime system on a selection of OpenMP applications, exhibiting both regular and irregular data reference patterns, resulting in average performance improvement of 28.1% on 8 processors.

Lecture Notes in Computer Science, 1996

This paper describes a performance tuning tool, named SVMview, for DSMbased parallel computers. SVMview is a tool for doing a post-mortem analysis of page movements generated during the execution of Fortran-S programs. This tool is able to analyze some particular phenomena such as false-sharing which occurs when several processors write to the same page simultaneously. Such behavior entails badly the performance of DSM systems, and thus are particularly important to be detected.

1997

Current parallelizing compilers for message-passing machines only support a limited class of data-parallel applications. One method for eliminating this restriction is to combine powerful shared-memory parallelizing compilers with software distributed-shared-memory (DSM) systems. We demonstrate such a system by combining the SUIF parallelizing compiler and the CVM software DSM. Innovations of the system include compiler-directed techniques that: 1) combine synchronization and parallelism information communication on parallel task invocation, 2) employ customized routines for evaluating reduction operations, and 3) select a hybrid update protocol that pre-sends data by flushing updates at barriers. For applications with sufficient granularity of parallelism, these optimizations yield very good speedups eight processors on an IBM SP-2 and DEC Alpha cluster, usually matching or exceeding the speedup of equivalent HPF and message-passing versions of each program. Based on our experimental results, we point out areas where additional compiler analysis and software DSM improvements can be used to achieve good performance on a broader range of applications.

1999

Distributed Shared-Memory (DSM) multiprocessors provide an attractive combination of cost-effective commodity architecture and, thanks to the shared-memory abstraction, relative ease of programming. Unfortunately, it is well known that tuning applications for scalable performance in these machines is time-consuming. To address this problem, programmers use performance monitoring tools. However, these tools are often costly to run, especially if highly-processed information is desired. In addition, they usually cannot be used to experiment with hypothetical architecture organizations.

In this paper, a Simple Distributed Shared Memory (SDSM) system, a hybrid version of shared memory and message passing version is proposed. This version effectively uses the benefits of shared memory in terms of easiness of programming and message passing in terms of efficiency. Further it is designed to effectively utilize the state-of-art multicore based network of workstations and supports standard pthread interface and OpenMP model for writing shared memory programs as well as MPI interface for writing message passing programs. This system has been studied using standard SPLASH-2 (Stanford ParalleL Applications for SHared memory -2), NPB(NAS Parallel Benchmarks), IMB (Intel MPI Benchmarks) benchmarks and some well known parallel algorithms. Its performance has been compared with JIAJIA DSM system that uses efficient scope consistency model for shared memory programs and with MPI library on network of Linux systems for MPI programs. In many cases it was found to perform much better than those systems.

2002

Belgrade Distributed shared memo y combines the advantages of sharedmemo y multiprocessors and distributed computer systems. Evaluations of four experimental orcommercial approaches to hardware DSMshow its potential for largescale, highperformance multiprocessor systems. istributed shared-memory systems are a successful hybrid of two parallel computer classes: shared-memory multiprocessors and distributed computer systems.l They provide the shared-memory abstraction in distributed computer systems. Thus, physically distributed memory becomes logically shared. DSM combines the advantages of both classes. The widely used sharedmemory programming model is simple, general, and efficient. Programs using t h~s model are highly portable. The application of distributed computing can readily acheve very powerful and cost-effective systems. These systems are claimed to be scalable because they can achieve an almost Linear performance increase with growing system size and investment (size-scalabhty). Because of these advantages, DSM is one of the most attractive solutions for building large-scale, high-performance multiprocessor systems. DSM systems vary greatly in regard to their architectures, algorithms,

International Parallel and Distributed Processing Symposium/International Parallel Processing Symposium, 2001

This work introduces a new technique that enables SDSMs to categorize dynamically and accurately memory sharing patterns in both classes of regular and irregular applications. The categorization is carried out automati- cally at run-time on a per-page basis, requiring no user or compiler assistance. We evaluate the potential benefits of our technique using execution-driven simulations of 8 applications running on

2002

Distributed Shared Memory (DSM) systems provide a logically shared memory over physically distributed memory to enable parallel computation on Networks of Workstations (NOWs). In this paper, we propose an infrastructure for DSM systems to utilize idle cycles in the network by thread migration.

2009

Understanding and optimizing the synchronization operations of parallel programs in distributed shared memory multiprocessors ( dsm), is one of the most important factors leading to significant reductions in execution time.