Most applications described on all these “nearby” web sites, forums, classes, tutorials, etc. involve relatively small thread-programs. Things rather closer to array-processing than to the kind of use I am exploring.
Imagine using GPU / CUDA / Tesla C 1060 for “Life”
Then imagine scaling the complexity of the rules and “organisms” 'way 'way up. And understand that the kinds of speed increases made available by CUDA for graphics is (possibly) irrelevant. (The code running for each automaton need not be more than a page or two, but each automaton would then be calling functions and accessing truly global memory all the time. Don’t be horrified! This the opposite end of the spectrum from array-processor / graphics use of parallel processing.)
Who out there is working on things like this? Can anybody refer me to literature or community members? and please forgive my ignorant newbie-ish-ness. My school’s CS dep’t just ordered our very first CUDA devices.
Please respond to my email as well as to the forum if it’s at all convenient. firstname.lastname@example.org
Dr. Chris Lanz
SUNY / Potsdam