It was an odd thought, but is forming into a real adventure in open source revenue. I think it is completely feasible to have Linux computers connected in the background to a composite net that computes things which cannot be done with even the largest super computers. And it seems to me that every person connected to that project should share evenly in the revenue that is generated. An easy example is a blender movie concept. If 100,000 people agree to be part of an animated movie, they download the free blender software and render a frame ( a few minutes ) selected from a site, then place that frame back on the site and the movie is distributed. I know that it is possible to devise a matrix computer strategy that outperforms a super computer.
I see this as a real possibility to apply the matrix style model to real problems if the software were written properly. The greatest cost in using super computers in my experience is the software and it's time to development. Ballistic Crays of the old days were very fast and cost 10M$, but the software cost much more than that and was only specific to that machine.
I would be interested in what others think about this possibility and how it might be used for the advantage of the Open source community ?
2 comments:
I think what you're proposing is that everybody run something equivalent to http://boinc.berkeley.edu/, and more software gets written to take advantage of that. I'm all for it!
I used to contribute some of my CPU power to the SETI@home project.
Yea, it sounds like a great idea to me too.
Post a Comment