Replicating Data for Better Performances in X10
نویسندگان
چکیده
Linguistic primitives for replica-aware coordination offer suitable solutions to the challenging problems of data distribution and locality in large-scale high performance computing. The data replication mechanisms that had previously been designed to extend Klaim with replicated tuples are now used to experiment with X10, a parallel programming language primarily targeting clusters of multi-core processors linked in a large-scale system via high-performance networks. Our approach aims at allowing the programmer to specify and coordinate the replication of shared data items by taking into account the desired consistency properties. The programmer can hence exploit such flexible mechanisms to adapt data distribution and locality to the needs of the application, in order to improve performance in terms of concurrency and data access. We investigate issues related to replica consistency and provide a performance analysis, which includes scenarios where replica-based specifications and relaxed consistency provide significant performance gains.
منابع مشابه
Substituent Effects on the Structural and Nonlinear Optical Properties of 1-[4-({(E)-[4-(methylsulfanyl)phenyl]methylidene}amino)phenyl]ethanone and Some of its Substituted Derivatives- a Theoretical Method
This work investigates the structural and nonlinear optical properties of a D-A type 1-[4-({(E)-[4-(methylsulfanyl)phenyl]methylidene}amino)phenyl]ethanone, MMP in which charge transfer occurs from -SCH3 donor to -COCH3 acceptor group through methylidene backbone; and some of its modeled analogues using quantum chemical calculations with pure BLYP and hybrid B3LYP correlation with high basis se...
متن کاملPrediction of RO Membrane Performances by Use of Adaptive Network-Based Fuzzy Interference Systems
This study aims to develop an Adaptive Network-based Fuzzy Inference System technique (ANFIS) and using the parameters of a complex mathematical model in the RO membrane performances. The ANFIS was constructed by using a subtractive clustering method to generate initial fuzzy inference systems. The model trained by 70% of the data set and then its validity is examined by remained 30% data set. ...
متن کاملDaily Pan Evaporation Modelling With ANFIS and NNARX
Evaporation, as a major component of the hydrologic cycle, plays a key role in water resources development and management in arid and semi-arid climatic regions. Although there are empirical formulas available, their performances are not all satisfactory due to the complicated nature of the evaporation process and the data availability. This paper explores evaporation estimation methods based o...
متن کاملImproving X10 Program Performances by Clock Removal
X10 is a promising recent parallel language designed specifically to address the challenges of productively programming a wide variety of target platforms. The sequential core of X10 is an object-oriented language in the Java family. This core is augmented by a few parallel constructs that create activities as a generalization of the well known fork/join model. Clocks are a generalization of th...
متن کاملTypes for X10 Clocks
X10 is a modern language built from the ground up to handle future parallel systems, from multicore machines to cluster configurations. We take a closer look at a pair of synchronisation mechanisms: finish and clocks. The former waits for the termination of parallel computations, the latter allow multiple concurrent activities to wait for each other at certain points in time. In order to better...
متن کامل