Performance evaluation of moment‐method codes on an intel iPSC/860 hypercube computer
Kenneth Klimkowski, Hao Ling
Abstract
Kenneth Klimkowski, Hao Ling
Abstract
Abstract The performance of a moment‐method code on a parallel computer is analyzed and evaluated. Algorithmic complexity costs are discussed in terms of matrix size and a parameter of matrix partitioning, subblock size. The measured computational times of the matrix fill operation and LU solver are analyzed in terms of scaled efficiencies. © 1993 John Wiley & Sons, Inc.
OpenAlex reports 4 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Abstract The performance of a moment‐method code on a parallel computer is analyzed and evaluated. Algorithmic complexity costs are discussed in terms of matrix size and a parameter of matrix partitioning, subblock size. The measured computational times of the matrix fill operation and LU solver are analyzed in terms of scaled efficiencies. © 1993 John Wiley & Sons, Inc.
Key concepts: Intel iPSC, Parallel computing, Solver, Hypercube, Moment (physics), Matrix (chemical analysis), Computer science, Computational science