Access our database of in-depth knowledge to learn about high-performance computing from experts in the field.

Tutorials about Writing Parallel Code

After you have your cluster up and running, you can use these links to information about what you need to know about parallel computing, porting existing parallel code, or writing new parallel code so you can run your own parallel code.

• Introductions to Parallel Computing:

  • Parallelization - learn about the issues involved in writing parallel code

  • Getting to Know MPI - get to know the fundamental calls of MPI and build on your knowledge

  • Parallel Zoology - compare and contrast parallel computing types: distributed, grid, and cluster computing and more

  • Parallel Paradigm - understand what, how, and why about programming paradigms for parallel computing

• Tutorials on Writing Parallel Code:

  • Parallel Knock - an exhibition of elemental message-passing in a simple parallel code

  • Parallel Adder - a tutorial on converting a single-processor code with independent work into a simple parallel code

  • Parallel Pascal's Triangle - a tutorial on parallelizing a propagation-style single-processor code, implementing a calculation requiring local communication

  • Parallel Circle Pi - a tutorial on creating a load balancing parallel code divisible into independent work

  • Parallel Life - a tutorial on parallelizing a propagation-style single-processor code, implementing a two-dimensional calculation requiring local communication

  • Visualization Message-Passing with MacMPI - a tutorial on using MacMPI's monitor window to understand, debug, and optimize parallel code

• Dauger Research authored the following on Apple's site:
  • Introduction to MPI Programming on Mac OS X - an Apple Developer Connection article introducing MPI

• and on the MacResearch site:
  • Multicore Eroding Moore's Law - an op/ed article questioning the industry's party line and highlighting the what multicore really means for the computer industry

• Cluster Software Development Kit - download make files, project files, and code examples for a variety of different compiler and language combinations. Learn by example, pick apart these files for your use, or substitute in your own code.

• Message-Passing Interfaces (MPIs) - find out about the message-passing libraries available for your parallel computing projects

• Compiling the MPIs - read instructions describing how to compile your code with the message-passing libraries

• Parallel Applications - try out others' parallel codes

• Visualization - take advantage of the communications visualization window present in MacMPI_X

• "Plug-and-Play" Cluster Computing
  High-Performance Computing for the Mainstream - "a must-read"

  This article was published in the special Cluster Computing issue (Mar/Apr 2005) of Computing in Science and Engineering (CiSE), a joint publication of American Institute of Physics (AIP) and Institute of Electrical and Electronics Engineers (IEEE). Dean Dauger and Viktor Decyk describe the technology design of Mac clusters, addressing why it was necessary to "reinvent" the cluster computer and how it is meant to benefit users of parallel computing.

• "Plug-and-Play" Cluster Computing - HPC Designed for the Mainstream Scientist
  Computational Science - ICCS 2005
  LNCS 3515, p. 84 ff.
  5th International Conference, Atlanta, GA, USA, May 22-25, 2005. Proceedings, Part I
  Springer-Verlag Berlin Heidelberg 2005

  Dauger Research presented this paper at the International Conference on Computational Science at Emory University in Alanta, Georgia, USA, on May 23, 2005. Dr. Dean Dauger and Dr. Viktor Decyk describe how the the technological design of Mac clusters maximizes the benefits of parallel computing for users.

• "Plug-and-Play" Cluster Computing using Mac OS X

  In this paper, accepted at the IEEE Cluster Computing 2003 conference, Dr. Dean Dauger and Dr. Viktor Decyk describe how the Mac clusters' ease-of-use maximizes the benefits of parallel computing for users.

• AppleSeed: A Parallel Macintosh Cluster for Scientific Computing
  in J. Plasma Fusion Res., Vol. 79, No. 8 (2003), 772-779
  by Viktor K. Decyk and Dean E. Dauger

• Numerically-Intensive "Plug-and-Play" Parallel Computing

  At the IEEE Cluster Computing 2001 coference, Dr. Dean Dauger and Dr. Viktor Decyk presented this paper describing much of their experience, success, and vision of this form of parallel computing.

• Plasma Physics Calculations on a Parallel Macintosh Cluster
  in Physical Scripta, Vol. T84, 85-88, 2000
  by Viktor K. Decyk, Dean E. Dauger, and Pieter R. Kokelaar