|
Livermore loops (also known as the Livermore Fortran kernels or LFK) is a benchmark for parallel computers. It was created by Francis H. McMahon from scientific source code run on computers at Lawrence Livermore National Laboratory. It consists of 24 do loops, some of which can be vectorized, and some of which cannot. The benchmark was published in 1986 in ''Livermore fortran kernels: A computer test of numerical performance range''.〔F. H. McMahon. Livermore fortran kernels: A computer test of numerical performance range. Technical Report UCRL-53745, Lawrence Livermore National Laboratory, Livermore, CA, December 1986. NTIS report #(DE87009360 ).〕 The Livermore loops were originally written in Fortran, but have since been ported to many programming languages. Each loop carries out a different mathematical kernel . Those kernels〔Xingfu Wu. Performance Evaluation, Prediction and Visualization of Parallel Systems. Springer, 1999. ISBN 0-7923-8462-8. Page 144.〕 are: * hydrodynamics fragment * incomplete Cholesky conjugate gradient * inner product * banded linear systems solution * tridiagonal linear systems solution * general linear recurrence equations * equation of state fragment * alternating direction implicit integration * integrate predictors * difference predictors * first sum * first difference * 2-D particle in a cell * 1-D particle in a cell * casual Fortran * Monte Carlo search * implicit conditional computation * 2-D explicit hydrodynamics fragment * general linear recurrence equations * discrete ordinates transport * matrix-matrix transport * Planckian distribution * 2-D implicit hydrodynamics fragment * location of a first array minimum. ==References== 〔 抄文引用元・出典: フリー百科事典『 ウィキペディア（Wikipedia）』 ■ウィキペディアで「Livermore loops」の詳細全文を読む スポンサード リンク
|