On the Design and Implementation of an Effective Prefetch Strategy for DSM Systems

Wang, Hsiao-Hsi; Li, Kuan‐Ching; Wang, Kuo-Jen; Lu, Ssu-Hsuan

doi:10.1007/s11227-006-5430-x

Cited by 4 publications

(14 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…History prefetching utilizes temporal locality while aggregate prefetching utilizes spatial locality. Yet another prefetching strategy, EPS (Effective Prefetch Strategy) [25,33,35], focuses on improving the prefetching hit rate and reducing the number of occurrences of the Waiting Synchronization Phenomenon (WSP, defined on Sect. 3.1) and of the Accumulated Waiting Phenomenon (AWP).…”

Section: Prefetching Strategiesmentioning

confidence: 99%

“…Essentially, EPS [25,33,35] is designed by adding a number of components in HPS that assist with prefetching when an application is executed. EPS includes three strategies that include Filtering Unnecessary Prefetches, Distributing Prefetch Overhead, and Load Balancing with Barrier Synchronization, as shown in Fig.…”

Section: Effective Prefetch Strategymentioning

confidence: 99%

“…The Effective Prefetch Strategy (EPS) proposed in [25,33,35] attempts to filter out unnecessary prefetches which would be made if the History Prefetching Strategy (HPS) was applied. In addition, EPS employs three different methods to improve performance over the original HPS in DSM systems: Filtering Unnecessary Prefetches, Distributing Prefetch Overhead, and Load Balancing with Barrier Synchronization.…”

Section: Introductionmentioning

confidence: 99%

“…JIAJIA is implemented with a lazy release consistency protocol and also supporting scope consistency to maintain data consistency. The global shared memory is distributed across the processors, where each processor acts as the home of a portion of the shared memory [34].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Design and Implementation of an Agent Home Scheme Strategy for Prefetch-Based DSM Systems

Wang

et al. 2008

Int J Parallel Prog

Self Cite

View full text Add to dashboard Cite

In recent years, cluster computing has been widely investigated and there is no doubt that it can provide a cost-effective computing infrastructure by aggregating computational power, communication, and storage resources. Moreover, it is also considered to be a very attractive platform for low-cost supercomputing. Distributed shared memory (DSM) systems utilize the physical memory of each computing node interconnected in a private network to form a global virtual shared memory. Since this global shared memory is distributed among the computing nodes, accessing the data located in remote computing nodes is an absolute necessity. However, this action will result in significant remote memory access latencies which are major sources of overhead in DSM systems. For these reasons, in order to increase overall system performance and decrease this overhead, a number of strategies have been devised. Prefetching is one such approach which can reduce latencies, although it always increases the workload in the home nodes. In this paper, we propose a scheme named Agent Home Scheme. Its most noticeable feature, when compared to other schemes, is that the agent home distributes the workloads of each computing nodes when sending data. By doing this, we can reduce not only the workload of the home nodes by balancing the 123 522 Int J Parallel Prog (2008) 36:521-542 workload for each node, but also the waiting time. Experimental results show that the proposed method can obtain about 20% higher performance than the original JIAJIA, about 18% more than History Prefetching Strategy (HPS), and about 10% higher than Effective Prefetch Strategy (EPS).

show abstract

Section: Prefetching Strategiesmentioning

confidence: 99%

Section: Effective Prefetch Strategymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Design and Implementation of an Agent Home Scheme Strategy for Prefetch-Based DSM Systems

Wang

et al. 2008

Int J Parallel Prog

Self Cite

View full text Add to dashboard Cite

show abstract

“…The usual data unit is a memory page or an object. Page-based prefetching has been well-studied in previous work including [1,2,3,4]. In this paper, we focus on object prefetching techniques.…”

Section: Introductionmentioning

confidence: 99%

Path-Analytic Distributed Object Prefetching

Luo¹,

Lam²,

Wang³

2009

2009 10th International Symposium on Pervasive Systems, Algorithms, and Networks

View full text Add to dashboard Cite

Abstract-This paper presents our studies on the connectivity between objects and traversal behavior over the access paths among objects in order to devise profitable prefetching policies for object-based distributed systems. We propose a profiling strategy that can classify classes and fields into a handful of generic types exhibiting distinctive and exploitable access patterns during the runtime. Based on the classifications, we propose an improved algorithm of object prefetching to select best candidates to prefetch under practical message size limits. We implement the methodology into our JESSICA2 distributed Java virtual machine and evaluate its effectiveness. Our experimental results show that our prefetching policies are able to eliminate over 93% cache coherence protocol messages and halve the execution time for fine-grained applications.

show abstract

Towards implementation of a novel scheme for data prefetching on distributed shared memory systems

Wang

et al. 2008

J Supercomput

Self Cite

View full text Add to dashboard Cite

High speed networks and rapidly improving microprocessor performance make the network of workstations an extremely important tool for parallel computing in order to speedup the execution of scientific applications. Shared memory is an attractive programming model for designing parallel and distributed applications, where the programmer can focus on algorithmic development rather than data partition and communication. Based on this important characteristic, the design of systems to provide the shared memory abstraction on physically distributed memory machines has been developed, known as Distributed Shared Memory (DSM). DSM is built using specific software to combine a number of computer hardware resources into one computing environment. Such an environment not only provides an easy way to execute parallel applications, but also combines available computational resources with the purpose of speeding up execution of these applications. DSM systems need to maintain data consistency in memory, which usually leads to communication overhead. Therefore, there exists a number of strategies that can be used to overcome this overhead issue and improve overall performance. Strategies as prefetching have been proven to show great performance in DSM systems, since they can reduce data access communication latencies from remote nodes. On the other hand, these strategies also transfer unnecessary prefetching pages to remote nodes. In this research paper, we focus on the access pattern during execution of a parallel application, and then analyze the data type and behavior of parallel applications. We propose an adaptive 112 H.-H. Wang et al.data classification scheme to improve prefetching strategy with the goal to improve overall performance. Adaptive data classification scheme classifies data according to the accessing sequence of pages, so that the home node uses past history access patterns of remote nodes to decide whether it needs to transfer related pages to remote nodes. From experimental results, we can observe that our proposed method can increase the accuracy of data access in effective prefetch strategy by reducing the number of page faults and misprefetching. Experimental results using our proposed classification scheme show a performance improvement of about 9-25% over the same benchmark applications running on top of an original JIAJIA DSM system.

show abstract

On the Design and Implementation of an Effective Prefetch Strategy for DSM Systems

Cited by 4 publications

References 14 publications

Design and Implementation of an Agent Home Scheme Strategy for Prefetch-Based DSM Systems

Design and Implementation of an Agent Home Scheme Strategy for Prefetch-Based DSM Systems

Path-Analytic Distributed Object Prefetching

Towards implementation of a novel scheme for data prefetching on distributed shared memory systems

Contact Info

Product

Resources

About