Thanks a lot for your help william! I think i'll try for an even higher one  if i get it i'll give you more than a weeks cpu cycles! (not too many though, gotta keep going with gimps :P). I found out today that apparently last years record was a 44k bit number!
question though  in "(a*4013#1) and (a*4013#+1)."  did you only use prime numbers for 'a'? Does that increase the probability of those expressions being prime too?
