parallel computing and future

Here is a previous thread that might be of interest to you:

What is Grid computing? 

Jacob,

The following websites may be useful to you:

http://www.llnl.gov/computing/tutorials/parallel_comp/ 

http://www-unix.mcs.anl.gov/dbpp/text/book.html 

http://www.parallelpython.com/

As far as I know prallel programming can be done in Fortran, C, C++ and python.  

V. Hegadekatte, University of Karlsruhe, Germany

I used MPI (with C++) for parallel programming, but am unaware of its recent development.

You can get some introductory information from Guru's blog http://compmatsci.wordpress.com/.  The links are here.

There are two ways to approach the problem of parallelizing a code.  The first is to go the hard way and do the parallelization yourself by writing MPI calls as required.  A more convenient approach is to use a framework that hides the MPI calls from you.  I prefer the latter because it makes my life easier in some ways and I don't have to worry about load balancing issues.  However, a framework constrains the freedom that you have in designing your code and can make things that are trivial in serial codes quite difficult to implement.

I good introductory book on MPI and parallel programming is 

George E.
Karniadakis and Robert M. Kirby, Parallel Scientific Computing in C++ and MPI , Cambridge University  Press, UK,628 pages, 2003.

There are numerous parallel processing frameworks that have been developed.  I use the Common Component Architecture idea that is implemented in the Uintah Computational Framework .  We use both C++ and legacy Fortran in our code.  There are Java based frameworks such as JPPF too.  If you are using structured grids then SAMRAI is a good framework to use.  If all you need is a parallel linear algebra solver then PETSc or Hypre are good.

Parallel computing technology has been growing with the Interent and the wide availability of computer clusters, or the supercomputer of poor people.  The MPI technology is sophisticated and there are available codes and software on the Internet.  I believe many top CAE software today are being extensively used in parallel mode.

Essential parallel computing resources can also be found on the Netlib (http://www.netlib.org).

Many languages can be utilized for parallel computing, but I believe the Fortran is still popular because the availability of many "legacy" codes.

I didn't know too much about this but I think Parallel computing is the use of two or more processors (cores, computers) in combination to solve a single problem.

Please describe this term so that we can understand it. WordPress development company