Parallel Processing | IBM Release 1.93 specs

2.6.1. Parallel Processing

[Under construction]

FIDO is a powerful and versatile simulation engine, which can compute simulations using inhomogeneous cubic grids on uniprocessors, symmetric multiprocessors (SMPs), and clusters tied together with TCP/IP. From the POEMS user’s point of view, SMPs act just like uniprocessors, except that an N-way SMP needs its work divided up into at least N chunks, and the load balancing is manual--usually it’s easy. This section discusses multiprocessor operations; technical details are in Appendix C.

9

Image 13

IBM Release 1.93 manual Parallel Processing

Contents

IBM T. J. Watson Research Center Yorktown Heights, NY Page IBM T. J. Watson Research Center Yorktown Heights, NY HOW Poems Works Using Poems Motivation Chapter Introduction Philosophy Optimization StructurePage Poems system organization Front-End Script poems.cmd Program OrganizationScript Operation Postprocessor Empost Fdtd Engine FIDO/TEMPEST Cluster Control Visualization System VIS5D Parallel Processing Poems Command-Line Options Command ReferenceGlobal Group Lambda FreqFunction Hosts Macdef Which means that the host’s predefined hostname is not used Print Macro SET RandomseedSimulator Title World GroupVerbose Boundary Basicstep Material GroupXrange Yrange Define Parameters epsReal epsImag muReal muImag Block Object GroupFAN Hollowbox GratingTiledplane 3DCURVE CurveCylinder Source Group Output Group Command Group Field Postprocess GroupCAD Webpage Flux FarfieldIntegral Modematch List MOVIE3D Movie Slice Dissipation Variables Optimize GroupGuess Store LimitPenalty Merit Parameters Schedule GroupRange Symmetry Computational Domain Objects Materials Perfectly-Matched Layers Plane Waves Page Beam Sources Merit Functions Optimization Phase uniformity across a plane 10 Optimized V antenna refractive Worked Example Optimizing a V AntennaPage Worked Example Glass Ridge Waveguide to Free Space Coupler Worked Example Doped Silica Waveguide ModePredefined Constants Confine Reserved Names Arithmetic Operators Predefined Mathematical Functions ABS Logical OperatorsAcos Acosh Ceil ATAN2COS Elintk 20. LN IntegralMAX MIN ROOT1D RandomRound Sign Material Parameter Functions Analytical Pupil FunctionsFlattop Startup and Steady State Tempest and General Fdtd InformationTime step Page Appendix A. V-Antenna Optimization Run Poems Input DIPOLE2I.PAR END Material END Object END World Subdomain ALL END Source END CommandEND Output END Optimize END Postprocess Amplex PhaseexPage Page Page END Tempest Input File DIPOLE2I.PAR.IN Written by Phil Hobbs Pages of pointsource statements omitted ALL Postprocessor orders DIPOLE2I.ORDERSDIPOLE2IEXI DIPOLE2IEXQ DIPOLE2IEYQ DIPOLE2IEYIDIPOLE2IEZI DIPOLE2IEZQ POSTPROC.1.NAME MiddlefluxArray Amplex POSTPROC.1.PARMSTRING Array DIPOLE2IPHASEEXPOSTPROC.2.PARMSTRING FF2 POSTPROC.6.NAME DIPOLE2IPXArray Poyntingz DIPOLE2IPZ POSTPROC.10.PARMSTRING POSTPROC.9.PARMSTRINGPOSTPROC.11.NAME Slice Indexn POSTPROC.13.COMPARISONDOMAIN Slice PoyntingzDIPOLE2IPZXY0.BMP POSTPROC.14.NAME DIPOLE2IPXZX0.BMP POSTPROC.16.COMPARISONDOMAINPOSTPROC.17.NAME Slice Amplex DIPOLE2IPHASEEXZX0.BMP DIPOLE2IPHASEEXXY0.BMPPOSTPROC.20.COMPARISONDOMAIN DIPOLE2IDISSZX0.BMP DIPOLE2IEXQZX0.BMP POSTPROC.24.COMPARISONDOMAIN Run Results DIPOLE2I.SIMPLEX Page Page Page Page Page Tempest patches Fdtd and Tempest Tempest limitations Advice common to all or most Fdtd programs Sample X11 Configuration Window System Configuration Release Notes Running Vis5DWish list Beta Release Limitations Page Page Emdenormal Emunderflow IndexMatlab Maxordersources 81 Maxpointsources