TreadMarks: shared memory computing on networks of workstations

Amza, Cristiana; Cox, Alan L.; Dwarkadas, Sandhya; Keleher, Peter J.; Hong, Li; Rajamony, Ram; Ye, Weimin; Zwaenepoel, Willy

doi:10.1109/2.485843

Cited by 630 publications

(349 citation statements)

References 9 publications

Supporting

Mentioning

344

Contrasting

Unclassified

Order By: Relevance

“…The first VC version (called V C d ) uses diffs to update views, while the second one (called V C h ) uses the home-based protocol to achieve view consistency. Their performance is compared with that of the LRC (Lazy Release Consistency) model implemented in TreadMarks [1]. We have tested our implementations on two clusters.…”

Section: Preliminary Experimental Resultsmentioning

confidence: 99%

“…We have implemented the VC model based on TreadMarks [1], which is a page-based DSM system. In TreadMarks, a diff is used to represent modifications on a page.…”

Section: Methodsmentioning

confidence: 99%

“…Distributed Shared Memory has become an active area of research in parallel and distributed computing with the goals of making DSM systems more convenient to program and more efficient to implement [17,8,16,6,4,3,1,10,11].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

View-Oriented Parallel Programming and View-Based Consistency

Huang

Purvis

Werstein

2004

Parallel and Distributed Computing: Applications and Technologies

View full text Add to dashboard Cite

show abstract

Section: Preliminary Experimental Resultsmentioning

confidence: 99%

“…We have implemented the VC model based on TreadMarks [1], which is a page-based DSM system. In TreadMarks, a diff is used to represent modifications on a page.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

View-Oriented Parallel Programming and View-Based Consistency

Huang

Purvis

Werstein

2004

Parallel and Distributed Computing: Applications and Technologies

View full text Add to dashboard Cite

show abstract

“…Parallel programs exhibit speed up and improve performance in a DSM environment but it uses dedicated computers, example DSM systems are Treadmarks, JUMP etc. [5,6 ] Grid computing optimizes existing unused computational power and is useful for parallel programs which can run independently, example EU Data Grid, NASA grid, BOINC, SLINC, SETI etc. [7,8] Both computing paradigms reduce hardware requirements and lower the installation and implementation costs.…”

Section: Motivationmentioning

confidence: 99%

Exploiting Communication Framework To Increase Usage Of SMIG Model Among Users

Shah¹,

Mahajan²

2012

SpecialIssue

View full text Add to dashboard Cite

Abstract-Today is the era of parallel and distributed computing models. Recent developments in DSM, Grids and DSM based Grids focus on high end computations of parallelized applications. We have built SMIG (Shared Memory Integrated with Grid) which amalgamates DSM and Grid computing paradigms as a part of our doctoral work. Literature citations have indicated a lacuna in how technological models are communicated to potential users to increase usage. Based on this lacuna we have identified the potential users and prepared a communication framework to disseminate SMIG information in order increase its usage. Hence in this paper we have compared various communication techniques used for disseminating DSM, Grid and DSM based Grid models as surveyed from literature. We have further designed and implemented a communication framework to percolate SMIG information to users. In the communication framework we have plugged in various tools for information dissemination and feedback (apart from those found in the survey) for promoting usage of technology among volunteers and application developers. These are included in the communication framework, namely arranging overview sessions, passing written documentation like presentations, installation handbook, FAQs, and also providing an opportunity to use SMIG model. The detailed responses received from the users after implementing the communication framework are encouraging and indicates that such a communication framework can be used for disseminating other technology developments to potential users. This will prove useful in today's dynamic world where technological developments are happening on a day to day basis.

show abstract

“…Similar abstractions of a shared memory are TreadMarks [1], HIVE [3], Munin [5] and Tempest and Typhoon [10]. The DVSA functionalities are simpler than those of these systems because, while these packages have been developed to support user applications, DVSA is application independent and its goal is the development of application dependent run time supports.…”

Section: Dvsamentioning

confidence: 99%

Two Layers Distributed Shared Memory

Baiardi

Guerri²,

Mori

et al.

High-Performance Computing and Networking

View full text Add to dashboard Cite

Abstract. This paper presents a methodology to design a distributed shared memory by decomposing it into two layers. An application independent layer supplies the basic functionalities to access shared structures and optimizes these functionalities according to the underlying architecture. On top of this layer, that can be seen as an application independent run time support, an application dependent layer defines the most suitable consistency model for the considered class of applications and it implements the most appropriate caching and prefetching strategies for the consistency model. To exemplify this methodology, we introduce DVSA, a package that implements the application independent layer and SHOB, an example of the second layer. SHOB defines a release consistency model for iterative numerical algorithms and it implements the corresponding caching and prefetching strategies. We present some experimental results of the methodology and discuss the performance of a uniform multigrid method developed through SHOB on a massively parallel architecture, the Meiko CS2, and on a cluster of workstations.

show abstract

TreadMarks: shared memory computing on networks of workstations

Cited by 630 publications

References 9 publications

View-Oriented Parallel Programming and View-Based Consistency

View-Oriented Parallel Programming and View-Based Consistency

Exploiting Communication Framework To Increase Usage Of SMIG Model Among Users

Two Layers Distributed Shared Memory

Contact Info

Product

Resources

About