petsc-3.7.1 2016-05-15
Report Typos and Errors

KSPPIPECGRR

Pipelined conjugate gradient method with automated residual replacements. This method has only a single non-blocking reduction per iteration, compared to 2 blocking for standard CG. The Many brnon-blocking reduction is overlapped by the matrix-vector product and preconditioner application. Many br

KSPPIPECGRR improves the robustness of KSPPIPECG by adding an automated residual replacement strategy. Many brTrue residual and auxiliary variables are computed explicitly in a number of dynamically determined Many briterations to reset accumulated rounding errors. Many br

See also KSPPIPECG, which is identical to KSPPIPECGRR without residual replacements. Many brSee also KSPPIPECR, where the reduction is only overlapped with the matrix-vector product. Many br

Many br

Notes

MPI configuration may be necessary for reductions to make asynchronous progress, which is important for Many brperformance of pipelined methods. See the FAQ on the PETSc website for details. Many br

Contributed by

Siegfried Cools, Universiteit Antwerpen, Many brEXA2CT European Project on EXascale Algorithms and Advanced Computational Techniques Many br

Reference

S. Cools, E.F. Yetkin, E. Agullo, L. Giraud, W. Vanroose, "Analysis of rounding error accumulation in the Many brconjugate gradients method to improve the maximal attainable accuracy of pipelined CG". Many brSubmitted to SIAM J. Sci. Comp., February 2016. Many br

See Also

KSPCreate(), KSPSetType(), KSPPIPECR, KSPGROPPCG, KSPPIPECG, KSPPGMRES, KSPCG, KSPCGUseSingleReduction()

Level:intermediate
Location:
src/ksp/ksp/impls/cg/pipecgrr/pipecgrr.c
Index of all KSP routines
Table of Contents for all manual pages
Index of all manual pages