(Excuse me, I posted yesterday this on google groups, but now I realized
this is a request of features not a discussion, so I moved the post into
issue page.)
I have HUGE databases in cowpatty format I created with "pyrit -e
ESSID -o NAME.cow export_cowpatty" and I use "pyrit -e ESSID -i
NAME.cow -r file.cap attack_cowpatty" to test it with captured
handshake.
To do attack_cowpatty needs long hours of computation, with my 4 cores
always at 99.9% (the PSKs/second are ~580.000 on BackTrak4, (see issue
#103 for PSKs/s on last r208 svn version)
Of course, one solution can be to split databases on several PCs to
speed up the cowpatty attack, but I wonder if there is another
solution as to use GPU to do cowpatty_attack: I mean, if GPU can do
better than my 4 cores CPU to do PSK/s generation, why it is not
possible to use GPU for cowpatty attack?
Or maybe I must do some optimization to cowpatty file (i.e. reorder
exported hash in some particular mode, etc)?
Original issue reported on code.google.com by pyrit.lo...@gmail.com on 26 Jan 2010 at 9:45
Original issue reported on code.google.com by
pyrit.lo...@gmail.com
on 26 Jan 2010 at 9:45