Compiz Fusion, Python, Parallel

October 21, 2007

So I finally got a working install on my computer by replacing the graphics card. So I decided to update from Feisty to Gutsy. I restarted X after my update and couldn’t log back in. I got a “KDM critical error” dialog box that told me to see the KDM error log, which actually didn’t tell me anything. It was a bit odd that when I restarted my computer, everything worked fine. So if you get that error, try a reboot.

In other news, has anyone looked at Python in Parallel? I was having a discussion with my advisor about parallel computing and he said FORTRAN was the way to go, but I think there have to be some Python projects in the works. I wonder if there are any benchmarks done comparing FORTRAN and Python. I looked on the web page but thought there may be even more relevant examples out there. I’m particularly interested in scientific applications, so if you do know of one, please drop me a comment.

According to this blog, Compiz Fusion 0.6.0 has been released. This is the first stable release of CF which is a good step for the CF community. Congrats everyone, maybe I’ll check out CF when there is an open driver for my video card. 🙂


  1. Threads are quite easy in Python, and at the same time quite limited, because of GIL. To make use of multicore/multiprocessor hardware, Parallel Python seems to be a viable alternative (http://www.parallelpython.com/).

  2. First of all, for scientific computing with Python, you definitely want to be using numpy/scipy, to have good performance.

    The most advanced work for parallel computing with Python is currently being done in ipython1, which is, unlike its names suggests, a framework for parallel computing/clustering with Python. Not much info is available on the webpage, but if you google, you should find some.

  3. My Data Structures professor is really interested in parallel computing, so I’ll be sure to show him Parallel Python. (I’m also going to set up a cluster over winter break, so this could be really useful) If you hear about any good comparisons between Python and FORTRAN, I’d love to hear them.

  4. The ipython project has been doing some great stuff on parallel python. Check out the following link for more information:


    Re: the merits of python versus Fortran for parallel computing. If your problem is embarrassingly parallel then using something like ipython to spread the work is going to be a good way to go.

    If your problem is not embarassingly parallel then you might need to be working at a lower level with a compiled language that has the ability to parallise code for you.

    However, my experience is that in the latter case you are much better off trying to rethink your solution and turn it into an embarrassingly parallel one.

  5. While we are going embarrassingly parallel, we are going to be going fast. The argument that I’ve heard is that there has been so many advances in MPI in FORTRAN that it will work better than going with Python (which has been less developed.)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: