Using accurate arithmetics to improve numerical reproducibility and stability in parallel applications

Journal of Supercomputing

Volumn 18, Issue 3, 2001, Pages 259-277

  He, Yun ^a   Ding, Chris H Q ^a  

^a UNIVERSITY OF CALIFORNIA   (United States)

Author keywords

Climate models; Distributed memory architecture; Double double precision arithmetic; Reproducibility; Self compensated summation

Indexed keywords

CLIMATOLOGY; COMPUTER SIMULATION; CONVERGENCE OF NUMERICAL METHODS; MATHEMATICAL MODELS; MATRIX ALGEBRA; VECTORS;

DISTRIBUTED MEMORY ARCHITECTURE; DOUBLE-DOUBLE PRECISION ARITHMETIC;

PARALLEL PROCESSING SYSTEMS;

EID: 0035276894     PISSN: 09208542     EISSN: None     Source Type: Journal    
DOI: 10.1023/A:1008153532043     Document Type: Article

References (26)

1. Multiprecision translation and execution of Fortran programs
2. A Fortran multiple precision arithmetic package
3. Design, implementation and testing of extended and mixed precision BLAS
4. A parallel climate data assimilation package
5. Data organization and I/O in an ocean circulation model
6. Atmospheric data assimilation on distributed parallel supercomputers
7. Design and performance of a scalable parallel community climate model
8. What every computer scientist should know about floating-point arithmetic
9. Iterative methods for solvong linear systems
10. The explicit free surface method in the GFDL modular ocean model
11. Computational design of NCAR community climate model
12. Arithmetic
13. The symmetric eigenvalue problem
14. Algorithms for arbitrary precision floating point arithmetic
15. Parallel ocean general circulation modeling