Posts: 17
Threads: 4
Joined: May 2010
05-25-2010, 12:57 PM
vBulletin md5(md5($pass).$salt) speed test:
WinXP 32bit,GeForce 9800 GTX/9800 GTX+, 511MB, 2000Mhz, 16MCU
Single hash - bruteforce:
IGHASHGPU v0.62: 185M/s
oclHashcat v0.19: 180M/s
EGB v1.6.1: 160M/s
Multi-Hash - (500 hashes)
EGB v1.6.1: 0,85M/s
oclHashcat v0.19: 0,33M/s
IGHASHGPU v0.62: NA
Posts: 5,185
Threads: 230
Joined: Apr 2010
05-25-2010, 01:10 PM
(This post was last modified: 05-25-2010, 01:23 PM by atom.)
oclHashcat does not cache 1st md5($pass) which is why it can handle more than 500 salts.
Posts: 17
Threads: 4
Joined: May 2010
(05-25-2010, 01:10 PM)atom Wrote: oclHashcat does not cache 1st md5($pass) which is why it can handle more than 500 salts.
Yes it can but there is serious problem with the speed decrease:
list with ~340 000 salted hashes after 3 minutes of bruteforce:
HashCat
Code:
Recovered.: 73/341183 hashes, 54/275276 salts
Speed/sec.: - plains, 50 words
Progress..: 9080/456976 (1.99%)
Running...: 00:00:03:00
Estimated.: 00:02:29:17
oclHashCat
Code:
Speed.GPU1: 0.00M/s
Speed.GPU*: 0.00M/s
Recovered.: 61/341185 Digests, 42/275278 Salts
Progress..: 0/456976 (0.00%)
Running...: 00:00:03:00
Estimated.: --:--:--:--
2 core CPU is faster than OC GTX9800+ ?!?
Posts: 5,185
Threads: 230
Joined: Apr 2010
its working perfectly..
if you have 180M/s in single-hash and you load 340k salts then the expected speed-drop is 180.000.000 / 340.000 = 529 words/sec. thats why it shows you 0M/s but actually its working.
and its working correctly, since 529 words/sec is faster than hashcat which shows you only 50 words words/sec.
the number of recovered passes also fit in, since oclHashcat word-generator works a bit other than hashcat word-generator. but on the end they both will have the same result.
Posts: 5,185
Threads: 230
Joined: Apr 2010
done in v0.20, thread closed
