DOLIB: Distributed Object Library
Eduardo F. D'Azevedo,Charles Romine +1 more
- 01 Dec 1995
- pp 750-755
TL;DR: In this article, the authors describe the use and implementation of DOLIB (Distributed Object Library), a library of routines that emulates global or virtual shared memory on Intel multiprocessor systems.
read more
Abstract: This report describes the use and implementation of DOLIB (Distributed Object Library), a library of routines that emulates global or virtual shared memory on Intel multiprocessor systems. Access to a distributed global array is through explicit calls to gather and scatter. Advantages of using DOLIB include: dynamic allocation and freeing of huge (gigabyte) distributed arrays, both C and FORTRAN callable interfaces, and the ability to mix shared-memory and message-passing programming models for ease of use and optimal performance. DOLIB is independent of language and compiler extensions and requires no special operating system support. DOLIB also supports automatic caching of read-only data for high performance. The virtual shared memory support provided in DOLIB is well suited for implementing Lagrangian particle tracking techniques. We have also used DOLIB to create DONIO (Distributed Object Network I/O Library), which obtains over a 10-fold improvement in disk I/O performance on the Intel Paragon.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Explicit Management of Memory Hierarchy
Jarek Nieplocha,Robert W. Harrison,Ian Foster +2 more
- 01 Jan 1997
TL;DR: The shared memory NUMA model implemented in the tools is described, extensions for wide area computing environments are discussed, and major applications of the tools are reviewed, which currently total over one million lines of code.
A Study of I/o in a pArallel Finite Element Groundwater Transport Code
David R. Mackay,G. Mahinthakumar,Ed D'Azevedo +2 more
- 01 Sep 1998
TL;DR: The findings suggest having a single processor collect data and per form large block-contiguous operations may be quite effi cient and portable for up to 32 processor configurations, and implementors of parallel I/O systems may consider some form of buffering as per formed in EDONIO to speed up such I/Os.
EDONIO: Extended distributed object network I/O library
E.F. D`Azevedo,C.H. Romine +1 more
- 01 Mar 1995
TL;DR: EDONIO is an enhanced version of DONIO optimized for the Intel Paragon Systems using the new M-ASYNC access mode and may show a ten-fold improvement in performance over the native NX I/O routines.
•Proceedings Article
Parallelization of a multiregion flow and transport code using software emulated global shared memory and high performance FORTRAN
E.F. D`Azevedo,Jin-Ping Gwo +1 more
- 01 Feb 1997
TL;DR: A suite of multiregion groundwater flow and solute transport codes that use Galerkin and Lagrangian- Eulerian finite element methods were parallelized to test the compatibility of a global shared memory emulation software with a High Performance FORTRAN (HPF) compiler, and to obtain performance characteristics and scalability of the parallel codes.