I'm trying to speed up MD5 search (brute force) a little. Why? Because it's interesting
I wanted to have a function that generates all possible words of lenght 'x' that can be built from a given charset...so I can't use fixed "loops" or similar.
I ended up with a version that is able to compare about 2.8 millions of hashes per second (on one core of my Opteron). My actual 'speed-up' is not concerned with the MD5 hashing (I took this from the openssl library, should be fine), but with the generation of all possible words of length Y from charset X.
Now I wonder if there's still a better way or some improvement possible? So it's not a real problem, but it would still be interesting to talk about it.
Because the maximum speed I get for "only word generation" is somewhere about 3-4 millions per second, that confuses me (thought that would be much more).
So I uploaded the code to:
Anybody interested is welcome to join the conversation