Name
Abstract | ||
|---|---|---|
Object-oriented software design and C++ implementation has proven to be a powerful, expressive, and extensible paradigm for
general computing, as evidenced by its explosive growth and near-universal embracement by the computing industry. Large, sophisticated,
and powerful object-oriented frameworks— groups of related libraries—have been implemented in C++ to address the complexities
of large-scale scientific computing, especially on parallel architectures. Experience with these frameworks is making the
case that these merits are equally realizable for scientific computing. The A++/P++, OVERTURE, POOMA, and DOE 2000 Scientific
Template Library frameworks have become established in the development of complex parallel numerical applications at Los Alamos
National Laboratory and elsewhere. These applications include multimaterial hydrodynamics, adaptive mesh refinement (AMR),
and modelling combustion and incompressible flows within the complex geometries associated with internal combustion engines;
each of these represents current work in ASCI and/or DOE Grand Challenge projects at LANL. The advantages of these frameworks
are two-fold: they provide significant simplification by encapsulating the complexities associated with the distribution of
data and parallel computation on that data, and they provide portability between serial and parallel environments. The scalability
of the object-oriented C++ approach to software engineering is such that as yet there appears to be no obstacle ot its use
in considerably larger and more complex future frameworks and applications.
|
Year | DOI | Venue |
|---|---|---|
1998 | 10.1007/3-540-49255-0_145 | ECOOP Workshops |
Keywords | Field | DocType |
array-class libraries,optimizing transformation system,scientific computing,object oriented,internal combustion engine,parallel computer,software engineering,incompressible flow,adaptive mesh refinement | Obstacle,Software design,Computer science,Theoretical computer science,Adaptive mesh refinement,Computational science,Software portability,Extensibility,Distributed computing,Scalability | Conference |
Volume | ISSN | ISBN |
1543 | 0302-9743 | 3-540-65460-7 |
Citations | PageRank | References |
6 | 0.74 | 1 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Kei Davis | 1 | 6 | 0.74 |
| Daniel J. Quinlan | 2 | 652 | 80.13 |