FreeTrade: Thanks for your input and it's nice to see other people working on that too. We may should exchange our results. I like to do the programming very low level for that purpose. While assembler is a little bit too freaky and time intense to code with, we are doing it on very low leveled c. P.e.: We don't use any strings at all yet and just std math operations besides the radix sort. Your mentioned SHA hashing doesn't make sense within the proof of work algorithm itself in my eyes. The created algo has to be as good as the well known hashing operations in terms of irreversibility and unique hashes anyway and then nothing else is needed around it. I think our current mechanism is already very good in terms of that.