MPI-2 Overview

14.01.02


Start Lecture by clicking here


Content

Author: Rolf Rabenseifner

  1. MPI-2 Overview
  2.     Acknowledgements
  3.     MPI-2 Outlook
  4.     MPI-1.2
  5.     MPI 1.2 — Clarifications to MPI 1.1
  6.     MPI 1.2 — Clarifications to MPI 1.1 (continued)
  7.     Dynamic Process Management
  8.     Dynamic Process Management — Get the intercomm, I.
  9.     Dynamic Process Management — Get the intercomm, II.
  10.     Dynamic Process Management — Multi-merging, a Challenge
  11.     Dynamic Process Management — MPI_Info Object
  12.     Dynamic Process Management — Establishing Communication
  13.     Dynamic Process Management — Another way
  14.     Dynamic Process Management — Singleton INIT
  15.     One-Sided Operations
  16.     One-Sided Operations — Origin and Target
  17.     One-Sided Operations — An Example
  18.     Extended Collective Operations
  19.     Extended Collective Operations — MPI_Bcast on intercomm.
  20.     Extended Collective Operations — MPI_Allgather on intercomm.
  21.     Extended Collective Operations — “In place” Buffer Specification
  22.     Extended Collective Operations — Generalized All-to-all
  23.     MPI - I/O
  24.     MPI - I/O — Logical view / Physical view
  25.     MPI - I/O — Definitions
  26.     MPI - I/O — Example with Subarray: Reads and distributes a matrix
  27.     Other MPI-2 Features (1)
  28.     Other MPI-2 Features (2)
  29.     Deprecated Names/Functions
  30.     Process Startup
  31.     C++ Language Bindings
  32.     Fortran 90 Support, I.
  33.     Problems Due to Data Copying and Sequence Association
  34.     Problems Due to Data Copying and Sequence Association
  35.     Fortran Problems with Register Optimization and 1-sided
  36.     Fortran 90 Support, II.
  37.     Summary of MPI-2
  38.     Other MPI Activities

Back to the Parallel Programming Workshop Overview