0.80 multiGPU alpha2 - newer and betterer! (ALPHA3 OUT, GO!)

Discussion and support for the CUDA Multiforcers (Windows and Linux)
  • Ads

0.80 multiGPU alpha2 - newer and betterer! (ALPHA3 OUT, GO!)

Postby Bitweasil » Wed Nov 24, 2010 6:59 am

Had some time tonight... fixed a few things...

- NTLM support!
- Some GPU speedups over 0.80 alpha 1 (minor, but present)
- Found and unfound hash file output.
- Pause! (press 'p' to pause)
- Better exit handling (Ctrl-C will now produce output files before exiting)
- Better default thread/block counts for most people

Details on the changes:
- NTLM support is present and seems to be fully working. So, you have both NTLM and MD5 to play with now. This, alone, should make it useful in production for some people.
- Still no Fermi tweaks, but things should be a bit faster for the same settings. I fixed some overhead.
- It now has proper "found" and "not found" hash output. Found is the same as before - '-o filename' - and it now puts just hash:password instead of the goofy stuff I was doing before. For the "not found" output, use '-n filename' - on exit, it will dump all hashes that have not been found.
- Currently, I still only support straight hash lists - pwdump/etc support is coming at some point soon here, but not here yet.
- You can pause the code! Press 'p' and it pauses pretty darn quick. Press any key to continue.
- Instead of ctrl-c just dropping out and losing the work, I now trap ctrl-c and exit in a controlled fashion. If you have output files, they should be written out as well as a summary of found hashes.
- The defaults are now: '--threads 512 --blocks 120 -m 100' - for modern cards, this is a MUCH better setup than what I had. If you have an older card or want less GUI interruption, you can override things as always.
- I'm also using the signal-based option for waiting on the GPUs. This *dramatically* reduces CPU usage during operation... leaving more time open for the CPUs to assist. ;-)
- Also, I used the LD_RUN_PATH environment variable properly, so you can just run it and it will find the stuff in lib64 automagically.


Questions? Comments?
Attachments
CUDA-Multiforcer-0_80_alpha2.tar.bz2
(563.83 KiB) Downloaded 471 times
Bitweasil
Site Admin
 
Posts: 912
Joined: Tue Jan 20, 2009 4:26 pm

Re: 0.80 multiGPU alpha2 - newer and betterer!

Postby Bitweasil » Wed Nov 24, 2010 7:20 am

On my 2x295 box, for NTLM:

Code: Select all
+------------------------------------------------------------------------------+
|'p' to pause          Cryptohaze Multiforcer 0.80 ALPHA2                      |
|                                                                              |
|Current PW len:         6 |      Passwords Found      |                       |
|Total hashes  :         6 |                           |  0: GPU: 690.53M/s    |
|Cracked hashes:         8 |  GoOd!                    |  1: GPU: 690.62M/s    |
|Total time    : 00:02:03  |  GPGPU                    |  2: GPU: 689.33M/s    |
|WUs: 68/172 (39.5%)       |  G0o|)                    |  3: GPU: 689.27M/s    |
|--------------------------|  Bwah                     |                       |
| Thread 3 mem loaded      |  1234                     |                       |
| Thread 2 mem loaded      |  P4s$                     |                       |
| Thread 1 mem loaded      |  abc                      |                       |
| Thread 0 mem loaded      |  bad                      |                       |
| Creating 4 threads       |                           |                       |
| Starting pw len 6        |                           |                       |
| Starting pw len 6        |                           |                       |
| Threads joined           |                           |                       |
| Td 0: out of WU.         |                           |                       |
| Td 1: out of WU.         |                           |                       |
| Td 2: out of WU.         |                           |                       |
| Thread 2 mem loaded      |                           |                       |
| Td 3: out of WU.         |                           |                       |
| Thread 3 mem loaded      |                           | TOTAL  : 2759.75M/s   |
+------------------------------------------------------------------------------+
Bitweasil
Site Admin
 
Posts: 912
Joined: Tue Jan 20, 2009 4:26 pm

Re: 0.80 multiGPU alpha2 - newer and betterer!

Postby blazer » Wed Nov 24, 2010 11:09 am

ok finally got it working

Spec: GTX470
Param: min 1 --max 6 -u charset-all-multi -l -t 512 -b 120

Multiforcer:


90000 Hashes
836 - MD5
Drops to 700 when it hits len 6

1054 - NTLM
Drops to 970 when it hits len 6

Single Hash
882 -MD5 Single
Drops to 800 when it hits len 6

1330 - NTLM Single
Drops to 1220 when it hits len 6

Ighash

90000 Hashes
MD5: 946
NTLM: 955

Single Hash
MD5:1478
NTLM: 2332 <- WOWOWOW

Two Hashes:

MD5: 1031
NTLM: 1462
blazer
 
Posts: 104
Joined: Fri Jan 23, 2009 10:18 am

Re: 0.80 multiGPU alpha2 - newer and betterer!

Postby Bitweasil » Wed Nov 24, 2010 2:26 pm

Thanks. I'm pretty sure they're doing hash unrolling, which is where that extra performance comes from. I'm definitely looking into that, just haven't implemented it yet.
Bitweasil
Site Admin
 
Posts: 912
Joined: Tue Jan 20, 2009 4:26 pm

Re: 0.80 multiGPU alpha2 - newer and betterer!

Postby hackajar » Wed Nov 24, 2010 11:18 pm

Code: Select all
./CUDA-Multiforcer-CPP -l -m 200 -b 768 -t 768 -h NTLM -f ntlm.txt --min=7 --max=7 -c charsets/charset-upper-lower-numeric-symbol-95.chr


Amazon EC2 Tesla M2050
Code: Select all
+------------------------------------------------------------------------------+
|'p' to pause          Cryptohaze Multiforcer 0.80 ALPHA2                      |
|                                                                              |
|Current PW len:         7 |      Passwords Found      |                       |
|Total hashes  :         1 |                           |  0: GPU: 1245.91M/s   |
|Cracked hashes:         0 |                           |  1: GPU: 1241.89M/s   |
|Total time    : 00:00:26  |                           |                       |
|WUs: 8/16260 (0.0%)       |                           |                       |
|--------------------------|                           |                       |
| Thread 0 mem loaded      |                           |                       |
| Thread 1 mem loaded      |                           |                       |
| Creating 2 threads       |                           |                       |
| Alloc 512MB bitmap       |                           |                       |
| Starting pw len 7        |                           |                       |
| Starting pw len 7        |                           |                       |
| Added GPU device 1       |                           |                       |
| Added GPU device 0       |                           |                       |
| chr isMulti: 1           |                           |                       |
|                          |                           |                       |
|                          |                           |                       |
|                          |                           |                       |
|                          |                           |                       |
|                          |                           | TOTAL  : 2487.80M/s   |
+------------------------------------------------------------------------------+

Core 2 Duo 2.4Ghz 8gig RAM 2x GTX480
Code: Select all
+------------------------------------------------------------------------------+
|'p' to pause          Cryptohaze Multiforcer 0.80 ALPHA2                      |
|                                                                              |
|Current PW len:         7 |      Passwords Found      |                       |
|Total hashes  :         1 |                           |  0: GPU: 1603.76M/s   |
|Cracked hashes:         0 |                           |  1: GPU: 1605.64M/s   |
|Total time    : 00:00:57  |                           |                       |
|WUs: 36/16260 (0.2%)      |                           |                       |
|--------------------------|                           |                       |
| Thread 1 mem loaded      |                           |                       |
| Thread 1 mem loaded      |                           |                       |
| Creating 2 threads       |                           |                       |
| Alloc 512MB bitmap       |                           |                       |
| Starting pw len 7        |                           |                       |
| Starting pw len 7        |                           |                       |
| Added GPU device 1       |                           |                       |
| Added GPU device 0       |                           |                       |
| chr isMulti: 1           |                           |                       |
|                          |                           |                       |
|                          |                           |                       |
|                          |                           |                       |
|                          |                           |                       |
|                          |                           | TOTAL  : 3209.39M/s   |
+------------------------------------------------------------------------------+
hackajar
 
Posts: 30
Joined: Tue Sep 14, 2010 10:21 pm

Re: 0.80 multiGPU alpha2 - newer and betterer!

Postby barbsie » Thu Nov 25, 2010 9:40 am

I have dual GTX480. The Cards are not detected in new version

Code: Select all
:~/gpu/CUDA-Multiforcer-0_80_alpha2$ ./CUDA-Multiforcer-CPP -l -m 500 -b 60 -t 512 -h NTLM -f test_hashes/Hashes-NTLM-Full.txt --min=7 --max=7 -c charsets/charset-upper-lower-numeric-symbol-95.chr
This program requires a CUDA-capable video card.
No cards found.  Sorry.  Exiting.


In older version, only 1 gets detected:

Code: Select all
~/gpu/CUDA-Multiforcer-Linux-0.72$ ./CUDA-Multiforcer -l -m 500 -b 60 -t 512 -h NTLM -f test_hashes/Hashes-NTLM-Full.txt --min=7 --max=7 -c charsets/charset-upper-lower-numeric-symbol-95.chr
Cryptohaze.com CUDA Multiforcer (multiple hash brute forcer)
by Bitweasil
Version 0.72, length 0-14
Currently supported hash types: MD5 FASTMD5 MD4 FASTMD4 NTLM FASTNTLM SHA1 FASTSHA1
Hash type: NTLM
CUDA Device Information:
Device 0: "GeForce GTX 480"
  Number of cores:                               120
  Clock rate:                                    1.40 GHz
  Performance Number:                            21015
  Note: Performance number is clock in mhz * core count, for comparing devices.
Single charset loaded.
Loading & sorting hashes.  This may take a while.
Hashes loaded (14 hashes)
You may want to consider the FAST[hash] option if it exists.
Launching kernel for password length 7
^C^C^C^C07%  Step rate: 1573.0M/s Search rate: 22022.2M/sec
barbsie
 
Posts: 1
Joined: Thu Nov 25, 2010 9:37 am

Re: 0.80 multiGPU alpha2 - newer and betterer!

Postby blazer » Thu Nov 25, 2010 10:12 am

update to the newer drivers i think 210. something (goto the nvidia website and download), you are probably using 195 or 197

old version only worked for 1 device anyway, thats why only one was detected
blazer
 
Posts: 104
Joined: Fri Jan 23, 2009 10:18 am

Re: 0.80 multiGPU alpha2 - newer and betterer!

Postby Bitweasil » Thu Nov 25, 2010 4:49 pm

barbsie wrote:I have dual GTX480. The Cards are not detected in new version


This code is built using the CUDA 3.2 SDK. You'll need the latest drivers (the 260 revision) for it to work. Stick them on & see if you still have the same issues.
Bitweasil
Site Admin
 
Posts: 912
Joined: Tue Jan 20, 2009 4:26 pm

Re: 0.80 multiGPU alpha2 - newer and betterer!

Postby someone0012 » Thu Nov 25, 2010 10:58 pm

Source code? Thanks.
someone0012
 
Posts: 3
Joined: Thu Nov 25, 2010 10:57 pm

Re: 0.80 multiGPU alpha2 - newer and betterer!

Postby Bitweasil » Thu Nov 25, 2010 11:16 pm

someone0012 wrote:Source code? Thanks.


Maybe, at some point, once it's done... why?
Bitweasil
Site Admin
 
Posts: 912
Joined: Tue Jan 20, 2009 4:26 pm

Next

Return to CUDA Multiforcers

Who is online

Users browsing this forum: No registered users and 1 guest

cron