The Experts below are selected from a list of 12606 Experts worldwide ranked by ideXlab platform

Michael L Scott - One of the best experts on this subject based on the ideXlab platform.

  • Conflict Detection and validation strategies for software transactional memory
    International Symposium on Distributed Computing, 2006
    Co-Authors: Michael Spear, Virendra J Marathe, William N Scherer, Michael L Scott
    Abstract:

    In a software transactional memory (STM) system, Conflict Detection is the problem of determining when two transactions cannot both safely commit. Validation is the related problem of ensuring that a transaction never views inconsistent data, which might potentially cause a doomed transaction to exhibit irreversible, externally visible side effects. Existing mechanisms for Conflict Detection vary greatly in their degree of speculation and their relative treatment of read-write and write-write Conflicts. Validation, for its part, appears to be a dominant factor—perhaps the dominant factor—in the cost of complex transactions. We present the most comprehensive study to date of Conflict Detection strategies, characterizing the tradeoffs among them and identifying the ones that perform the best for various types of workload. In the process we introduce a lightweight heuristic mechanism—the global commit counter—that can greatly reduce the cost of validation and of single-threaded execution. The heuristic also allows us to experiment with mixed invalidation, a more opportunistic interleaving of reading and writing transactions. Experimental results on a 16-processor SunFire machine running our RSTM system indicate that the choice of Conflict Detection strategy can have a dramatic impact on performance, and that the best choice is workload dependent. In workloads whose transactions rarely Conflict, the commit counter does little to help (and can even hurt) performance. For less scalable applications, however—those in which STM performance has traditionally been most problematic—it can improve transaction throughput many fold.

  • Conflict Detection and validation strategies for software tradsacfional memory
    Lecture Notes in Computer Science, 2006
    Co-Authors: Michael Spear, Virendra J Marathe, William N Scherer, Michael L Scott
    Abstract:

    In a software transactional memory (STM) system, Conflict Detection is the problem of determining when two transactions cannot both safely commit. Validation is the related problem of ensuring that a transaction never views inconsistent data, which might potentially cause a doomed transaction to exhibit irreversible, externally visible side effects. Existing mechanisms for Conflict Detection vary greatly in their degree of speculation and their relative treatment of read-write and write-write Conflicts. Validation, for its part, appears to be a dominant factor-perhaps the dominant factor-in the cost of complex transactions. We present the most comprehensive study to date of Conflict Detection strategies, characterizing the tradeoffs among them and identifying the ones that perform the best for various types of workload. In the process we introduce a lightweight heuristic mechanism-the global commit counter-that can greatly reduce the cost of validation and of single-threaded execution. The heuristic also allows us to experiment with mixed invalidation, a more opportunistic interleaving of reading and writing transactions. Experimental results on a 16-processor SunFire machine running our RSTM system indicate that the choice of Conflict Detection strategy can have a dramatic impact on performance, and that the best choice is workload dependent. In workloads whose transactions rarely Conflict, the commit counter does little to help (and can even hurt) performance. For less scalable applications, however-those in which STM performance has traditionally been most problematic-it can improve transaction throughput many fold.

  • DISC - Conflict Detection and validation strategies for software transactional memory
    Lecture Notes in Computer Science, 2006
    Co-Authors: Michael Spear, Virendra J Marathe, William N Scherer, Michael L Scott
    Abstract:

    In a software transactional memory (STM) system, Conflict Detection is the problem of determining when two transactions cannot both safely commit. Validation is the related problem of ensuring that a transaction never views inconsistent data, which might potentially cause a doomed transaction to exhibit irreversible, externally visible side effects. Existing mechanisms for Conflict Detection vary greatly in their degree of speculation and their relative treatment of read-write and write-write Conflicts. Validation, for its part, appears to be a dominant factor—perhaps the dominant factor—in the cost of complex transactions. We present the most comprehensive study to date of Conflict Detection strategies, characterizing the tradeoffs among them and identifying the ones that perform the best for various types of workload. In the process we introduce a lightweight heuristic mechanism—the global commit counter—that can greatly reduce the cost of validation and of single-threaded execution. The heuristic also allows us to experiment with mixed invalidation, a more opportunistic interleaving of reading and writing transactions. Experimental results on a 16-processor SunFire machine running our RSTM system indicate that the choice of Conflict Detection strategy can have a dramatic impact on performance, and that the best choice is workload dependent. In workloads whose transactions rarely Conflict, the commit counter does little to help (and can even hurt) performance. For less scalable applications, however—those in which STM performance has traditionally been most problematic—it can improve transaction throughput many fold.

Chengzheng Sun - One of the best experts on this subject based on the ideXlab platform.

  • CSCW - Dependency-Conflict Detection in real-time collaborative 3D design systems
    Proceedings of the 2013 conference on Computer supported cooperative work - CSCW '13, 2013
    Co-Authors: Agustina, Chengzheng Sun
    Abstract:

    One core issue in real-time collaborative 3D design systems is dependency-Conflict caused by concurrent access to object replicas in Dependency Graph (DG) states -- the common underlying structure of 3D design systems. Prior work has provided an Operational Transformation (OT) solution to dependency-Conflict resolution, assuming dependency-Conflict can be detected. But dependency-Conflict Detection is a complex and challenging issue because dependency relations among targets of concurrent operations -- a necessary condition for dependency-Conflict -- is dynamic in nature and the Detection must refer to prior DG states that are often unavailable during dependency-Conflict Detection. Conflict Detection was never a difficult issue in prior OT work based on well-defined linear data models because Conflict relations are readily derivable from operation position parameters without having to refer to prior document states. This work is the first to recognize and solve the dependency-Conflict Detection problem, which complements prior work in dependency-Conflict resolution to provide a complete solution to dependency-Conflict problem in real-time collaborative DG-based systems. Both theoretical verification and real implementation have been applied to validate the proposed dependency-Conflict Detection solution.

  • CTS - Agile semantic Conflict Detection in real-time collaborative systems
    2009 International Symposium on Collaborative Technologies and Systems, 2009
    Co-Authors: Yang Zheng, Haifeng Shen, Chengzheng Sun
    Abstract:

    Real-time collaborative systems allow geographically dispersed users to concurrently work on shared artifacts. Conflict Detection and management present major challenges in the design of those systems. In this paper, we propose a novel technique for detecting “combine-related” semantic Conflicts in real-time collaborative systems. It can maximally capture the semantics of Conflicting “combine-based” operations and help users analyze Conflicts from different perspectives. The technique can provide maximal agility in detecting semantic Conflicts among operations with intricate interrelations. It can also significantly improve users' understanding of the Conflicts towards finding suitable solutions. The technique is particularly useful for Conflict Detection and resolution in semantic-rich collaborative systems such as word processing systems, CAD systems, image processing systems and digital media design systems.

Jin Weidong - One of the best experts on this subject based on the ideXlab platform.

  • Conflict Detection algorithm in collaborative design
    Journal of Computer Applications, 2007
    Co-Authors: Jin Weidong
    Abstract:

    Conflict is the essence of collaborative design. The causes of Conflicts include different thinking angles, judging criteria and specialty knowledge of different designers in collaborative design environment. A consistency model of operation sequence and Conflict Detection algorithm based on geometry level were proposed. Accordingly, a prototype Conflict Detection system in collaborative design was implemented.

Michael Spear - One of the best experts on this subject based on the ideXlab platform.

  • Conflict Detection and validation strategies for software transactional memory
    International Symposium on Distributed Computing, 2006
    Co-Authors: Michael Spear, Virendra J Marathe, William N Scherer, Michael L Scott
    Abstract:

    In a software transactional memory (STM) system, Conflict Detection is the problem of determining when two transactions cannot both safely commit. Validation is the related problem of ensuring that a transaction never views inconsistent data, which might potentially cause a doomed transaction to exhibit irreversible, externally visible side effects. Existing mechanisms for Conflict Detection vary greatly in their degree of speculation and their relative treatment of read-write and write-write Conflicts. Validation, for its part, appears to be a dominant factor—perhaps the dominant factor—in the cost of complex transactions. We present the most comprehensive study to date of Conflict Detection strategies, characterizing the tradeoffs among them and identifying the ones that perform the best for various types of workload. In the process we introduce a lightweight heuristic mechanism—the global commit counter—that can greatly reduce the cost of validation and of single-threaded execution. The heuristic also allows us to experiment with mixed invalidation, a more opportunistic interleaving of reading and writing transactions. Experimental results on a 16-processor SunFire machine running our RSTM system indicate that the choice of Conflict Detection strategy can have a dramatic impact on performance, and that the best choice is workload dependent. In workloads whose transactions rarely Conflict, the commit counter does little to help (and can even hurt) performance. For less scalable applications, however—those in which STM performance has traditionally been most problematic—it can improve transaction throughput many fold.

  • Conflict Detection and validation strategies for software tradsacfional memory
    Lecture Notes in Computer Science, 2006
    Co-Authors: Michael Spear, Virendra J Marathe, William N Scherer, Michael L Scott
    Abstract:

    In a software transactional memory (STM) system, Conflict Detection is the problem of determining when two transactions cannot both safely commit. Validation is the related problem of ensuring that a transaction never views inconsistent data, which might potentially cause a doomed transaction to exhibit irreversible, externally visible side effects. Existing mechanisms for Conflict Detection vary greatly in their degree of speculation and their relative treatment of read-write and write-write Conflicts. Validation, for its part, appears to be a dominant factor-perhaps the dominant factor-in the cost of complex transactions. We present the most comprehensive study to date of Conflict Detection strategies, characterizing the tradeoffs among them and identifying the ones that perform the best for various types of workload. In the process we introduce a lightweight heuristic mechanism-the global commit counter-that can greatly reduce the cost of validation and of single-threaded execution. The heuristic also allows us to experiment with mixed invalidation, a more opportunistic interleaving of reading and writing transactions. Experimental results on a 16-processor SunFire machine running our RSTM system indicate that the choice of Conflict Detection strategy can have a dramatic impact on performance, and that the best choice is workload dependent. In workloads whose transactions rarely Conflict, the commit counter does little to help (and can even hurt) performance. For less scalable applications, however-those in which STM performance has traditionally been most problematic-it can improve transaction throughput many fold.

  • DISC - Conflict Detection and validation strategies for software transactional memory
    Lecture Notes in Computer Science, 2006
    Co-Authors: Michael Spear, Virendra J Marathe, William N Scherer, Michael L Scott
    Abstract:

    In a software transactional memory (STM) system, Conflict Detection is the problem of determining when two transactions cannot both safely commit. Validation is the related problem of ensuring that a transaction never views inconsistent data, which might potentially cause a doomed transaction to exhibit irreversible, externally visible side effects. Existing mechanisms for Conflict Detection vary greatly in their degree of speculation and their relative treatment of read-write and write-write Conflicts. Validation, for its part, appears to be a dominant factor—perhaps the dominant factor—in the cost of complex transactions. We present the most comprehensive study to date of Conflict Detection strategies, characterizing the tradeoffs among them and identifying the ones that perform the best for various types of workload. In the process we introduce a lightweight heuristic mechanism—the global commit counter—that can greatly reduce the cost of validation and of single-threaded execution. The heuristic also allows us to experiment with mixed invalidation, a more opportunistic interleaving of reading and writing transactions. Experimental results on a 16-processor SunFire machine running our RSTM system indicate that the choice of Conflict Detection strategy can have a dramatic impact on performance, and that the best choice is workload dependent. In workloads whose transactions rarely Conflict, the commit counter does little to help (and can even hurt) performance. For less scalable applications, however—those in which STM performance has traditionally been most problematic—it can improve transaction throughput many fold.

Peng Jun - One of the best experts on this subject based on the ideXlab platform.