Compiler transformations for high-performance computing
ACM Computing Surveys (CSUR)
The AppLeS parameter sweep template: user-level middleware for the grid
Proceedings of the 2000 ACM/IEEE conference on Supercomputing
Data Reuse Analysis Technique for Software-Controlled Memory Hierarchies
Proceedings of the conference on Design, automation and test in Europe - Volume 1
A data distributed parallel algorithm for nonrigid image registration
Parallel Computing
The Anatomy of the Grid: Enabling Scalable Virtual Organizations
International Journal of High Performance Computing Applications
Virtual data Grid middleware services for data-intensive science: Research Articles
Concurrency and Computation: Practice & Experience - Middleware for Grid Computing
Pegasus: A framework for mapping complex scientific workflows onto distributed systems
Scientific Programming
Design and implementation of intelligent scheduler for Gaussian portal on quantum chemistry grid
ICCS'03 Proceedings of the 2003 international conference on Computational science: PartIII
Exploiting replication and data reuse to efficiently schedule data-intensive applications on grids
JSSPP'04 Proceedings of the 10th international conference on Job Scheduling Strategies for Parallel Processing
Combining performance aspects of irregular gauss-seidel via sparse tiling
LCPC'02 Proceedings of the 15th international conference on Languages and Compilers for Parallel Computing
A framework for the design and reuse of grid workflows
SAG'04 Proceedings of the First international conference on Scientific Applications of Grid Computing
Hi-index | 0.00 |
This paper presents a data minimization method that aims at reducing overhead for data reuse in grid environments. The data reuse here is designed to promote efficient use of grid resources by avoiding multiple executions of the same computation in a collaborative community. To promote this at the program block level, our method minimizes the data size of attribute values, which are used for identification of computation products stored in a database (DB) server. Because attribute values are specified in queries used for store, search, or retrieval of computation products, their reduction leads to less communication between computing nodes and the DB server, minimizing the runtime overhead of data reuse. We also show some experimental results obtained using a time-consuming medical application. We find that the method successfully reduces the data size of a query from 683 MB to 52 B. This reduction allows our data reuse framework to reduce execution time from approximately 9 minutes to 27 seconds.