A million little bugs - hashcat dev

atom is insane! He managed to get some optimizations to tune up the speed of attacks under 8 chars to the "fullspeed".

As a conclusion oclhashcat-lite takes now about 45 seconds for the whole space of 6 chars on a 5970 at stock clock in NTLM!

He also finished his implementation of DCC and got a very nice performance improve of about 36% to 4608.1M/s from 3420M/s in oclhashcat v0.25.

To give that some extra spice he also finished increment mode support on salted hashes.

Got some new informations about the default attack of oclhashcat-lite:

Since the unique features of oclhashcat is the per position charsets it will automaticly use it in the default attack of the lite version. The masks are set like this:
?1 = "?u?l"
?2 = "?l?d@ "
mask = "?1?l?l?l?2?2?2?2?2"

This will produce output like this:
"Abcdx1 2@"
"Tommy 123"

Another new feature will be a "current"-line in the status output:

Status....: Running
Length....: 8
Mask......: ?1?2?2?2?2?2?2?3
Current...: **2jkisc
Speed.GPU*: 9500.3M/s
Progress..: 333692928000/5668341202944 (5.89%)
Running...: 35 secs
Estimated.: 9 mins, 21 secs
[s]tatus [p]ause [r]esume [q]uit =>

oclHashcat lite v0.01 = 10.1B on hd5970@770mhz

Atom decided to take the increment option as default. So the previous post has to be adjusted.

The default attack of oclhashcat-lite will adress 5-8 chars!