Computational Technology Resources - CCP

C.R. Calidonna and M. Mango Furnari

Institute of Cybernetics, Italian National Research Council, Arco Felice, Naples, Italy

doi:10.4203/ccp.52.2.1

C.R. Calidonna, M. Mango Furnari, "Data Access Collection and Data Partitioning for NUMA Architectures", in B.H.V. Topping, (Editor), "Advances in Computational Mechanics with High Performance Computing", Civil-Comp Press, Edinburgh, UK, pp 33-40, 1998. doi:10.4203/ccp.52.2.1

Abstract

In this paper we propose an unified approach to manage the problems of data dependence analysis, unimodular loop transformation and data partitioning. This approach is based on the notion of data access descriptor, which captures the way how a data structure is effectively accessed in the different regions of a programs. We show that in our approach the previous problems could benefit of reusing the same kind of information, saving compiling time. Furthermore, the data access information could also be used to easily implement a test for the existence of communication free partition. A description of a new compiler phase, that collects data access functions, is described besides a brief description of the test above which need previous captured data information.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £56 +P&P)

	Computational & Technology Resources an online resource for computational, engineering & technology publications
	not logged in - login
Front Page Browse CCP CSETS CTR IJRT Other Authors Search Purchase Guide FAQ Contact us	Civil-Comp Proceedings ISSN 1759-3433 CCP: 52 ADVANCES IN COMPUTATIONAL MECHANICS WITH HIGH PERFORMANCE COMPUTING Edited by: B.H.V. Topping Paper II.1 Data Access Collection and Data Partitioning for NUMA Architectures C.R. Calidonna and M. Mango Furnari Institute of Cybernetics, Italian National Research Council, Arco Felice, Naples, Italy doi:10.4203/ccp.52.2.1 purchase the full-text of this paper Full Bibliographic Reference for this paper C.R. Calidonna, M. Mango Furnari, "Data Access Collection and Data Partitioning for NUMA Architectures", in B.H.V. Topping, (Editor), "Advances in Computational Mechanics with High Performance Computing", Civil-Comp Press, Edinburgh, UK, pp 33-40, 1998. doi:10.4203/ccp.52.2.1 Abstract In this paper we propose an unified approach to manage the problems of data dependence analysis, unimodular loop transformation and data partitioning. This approach is based on the notion of data access descriptor, which captures the way how a data structure is effectively accessed in the different regions of a programs. We show that in our approach the previous problems could benefit of reusing the same kind of information, saving compiling time. Furthermore, the data access information could also be used to easily implement a test for the existence of communication free partition. A description of a new compiler phase, that collects data access functions, is described besides a brief description of the test above which need previous captured data information. purchase the full-text of this paper (price £20) go to the previous paper go to the next paper return to the table of contents return to the book description purchase this book (price £56 +P&P)
Back to top	©Civil-Comp Limited 2023 - terms & conditions