| This takes Jacobi Iteration for Laplace's Equation in a 2D square and uses this to illustrate: |
| Programming in both Data Parallel (HPF) and Message Passing (MPI and a simplified Syntax) |
| SPMD -- Single Program Multiple Data -- Programming Model |
| Stencil dependence of Parallel Program and use of Guard Rings |
| Collective Communication |
| Basic Speed Up,Efficiency and Performance Analysis with edge over area dependence and consideration of load imbalance and communication overhead effects. |
001 CPS615 -- Base Course for the Simulation Track of Computational
Science
Fall Semester 1997 --
The Laplace Example
Programming and Performance
002 Abstract of Laplace Example for CPS615
003 Parallel Computing
Algorithms and
Software --
Laplace Example
004 The Solution of Laplace's Equation
005 Discretized Form of Laplace'e Equation on a Parallel Processor
006 Basic Structure of Domain to be Updated in Parallel Version
007 Sequential and Introduction to Parallel Coding for the
Laplace Example
008 SEQUENTIAL LAPLACE PROGRAMMING
JACOBI ITERATION IN ONE DIMENSION
(constant in y direction)
009 SEQUENTIAL LAPLACE PROGRAMMING
JACOBI ITERATION IN TWO DIMENSIONS
010 Approaches to Parallel Programming
011 SPMD or SCMD
Single Program (code) Multiple Data
012 Data Parallel
Programming for
Laplace Example
013 Parallel Laplace Programming
Data Parallel for Jacobi Iteration in One Dimension
014 Notes on HPF Implementation of Lapace Solver
015 Message Passing Model
Used for Parallel
Programming for
Laplace Example
016 Basic Message Passing Approach
017 Various Styles/Terms in Messaging Passing - I
018 Various Styles/Terms in Messaging Passing - II
019 Parallel Laplace Programming: Set Up of
Message Passing for Jacobi Iteration in One Dimension
020 Node Program: Message Passing for Laplace Sover
021 Collective Communication Primitives
022 Implementation of MPSHIFT(+1, SOURCE,DEST)
023 Possible Implementation of MPSHIFT in MPI
024 Implementation of SHIFT in MPI
025 Implementation of GLOBALMAX (TEST)
026 General Features
of
Laplace Example
027 What does the Laplace Update calculation look like?
028 The Various Stencil Special Cases
029 Communication Loads
030 What is the relation of sequential and parallel programming models
?
031 More on SPMD Programming Model and Sequential/Parallel Comparison
032 Programming with Guard Rings - Sequential
033 Programming with Guard Rings - Parallel
034 Special Case of Corner Processor
035 Analysis of Parallel Overheads
Efficiency and
Speedup
036 General Formalism for Speed Up
037 What Affects Speed Up ?
038 Load Imbalance and Speed-Up for Laplace Example -- I
039 Load Imbalance and Speed-Up for Laplace Example -- II
040 Analytical Analysis of Load Imbalance
041 Communication Overhead
042 Analytical Form of Speed Up for Communication Overhead
043 General Form of Efficiency
044 Communication to Calculation Ratio as a function of template
045 Performance for Increasing Stencil
046 Matrix Multiplication on the Hypercube
047 Efficiency of QCD Physics Simulation on JPL MarkIIIfp Hypercube
048 General Analysis of Overheads and Efficiency
049 Speed Up as a Function of Grain Size