Maximum PC

Against my better judgement, the other day I decided to take another bite out of the Project Euler apple and began working on Problem 391. I say against my better judgement because I'm stuck and desperately want to solve this damn problem! It didn't take long for me to find a O(n) solution that correctly solves the example of M(20), but unfortunately, it is nowhere near fast enough to solve a graph with 2^10000 nodes (or even 2^100, not that it matters). Not to mention the 2^9,999, 2^9,998, ... 2^1 other graphs that have to be solved. Yes, those stinking mathematicians have found one of those things they call a pattern (or series).

Anyways, I just wanted to wanted to see if anyone else is either working or interested in working on this problem. I've very curious to discover what method(s) they've used to cut such an enormous problem space down to size.



PS - Maybe I'm losing my memory, but I would swear that people were using signatures recently, no?