Primefinder, Part 2?
So, I’m tempted to revisit my brute-force primefinder, now that I’ve discovered that I can buy time on a supercomputer* for next to nothing. I looked into Amazon’s EC2 cloud a couple of years ago for a project I was doing for a client, found it wasn’t a good fit, and then never really thought about it any more. However, recently I started a project for another client that *is* a good fit, and as a result I’ve been putting EC2 through it’s paces – and I can see that doing things like brute-force testing a large problem space for potential solutions to the p/q problem is something EC2 would be *very* good at. (Any problem that can be broken into small chunks and would therefore run well on a cluster). And, CPU on EC2 is relatively cheap. Storage and bandwidth cost money, but we don’t really need either of those.
I know – just what I needed, another pet project. 😉 [actually, revisiting an old one]
* = well, very large cluster anyway