Thinking Parallel

A Blog on Parallel Programming and Concurrency by Michael Suess

News for Week 06/2007

A CameraIt’s this time of the month again when I am posting stuff I found interesting. I thought this was the most boring part of my blog, but my colleagues have told me that they liked these newsposts a lot, so I am keeping them up. Of course, I am also interesting in hearing what you think about the issue, would you like more news from other sites quoted here? Or less? Or none at all? Please let me know and leave a comment! Or write me personally, I always enjoy reading mail from my readers! But now without further ado, here is my list of newsworthy resources on the net:

  • David Chisnall is writing about Parallel Programming with POSIX here and here. Interestingly, he does not cover POSIX Threads, but sticks to processes instead. I also found the following quote refreshing: “Writing parallel code isn’t much more difficult than writing serial code.” – When you listen to many other people that are doing parallel programming, you sometimes get the feeling that it must be some alien-skill from out of space that nobody in their right mind can master. I don’t think so, though and obviously at least some people agree.
  • Steve Apiki is taking a look at parallel programming from a project managers perspective here, here and here. This is part of an ongoing series, will be interesting to see where he is heading.
  • Robert O’Callahan takes a first look at Fortress in his weblog.
  • Not really new, but not as old as some of the other articles I have declared as news is this article from Bruce Dawson about his development experiences with multiple cores on the Xbox 360. Very interesting for all game programmers out there who are thinking about going mult-threaded, especially since he points out what parts of a game-engine can be outsourced to different threads and how.
  • Patrick Logan warns, cautions and preaches against starting to use Transactional Memory. The article is a bit difficult to read, because he starts with some comments at the top. Its main argument appears to be that shared memory is too difficult for normal programmers and message passing is the answer. I have been meaning to write about this thesis for some time, but have not gotten around to doing it (yet).
  • There is a paper comparing the language support for generic programming in different programming languages available – interesting, if you are into generic programming.

Comments are closed