DES Cracker BETA

[Rolf]

I dont know, but maybe Михаил knows?

[schwarzwaldhacker]

Михаил, как дела, "Compiling Fermi": Как это сделать?

[IvanG]

Поставить CUDA SDK 3.0/3.1. И закомпилить .

[Rolf]

Как всё просто!
Schwarzwaldhacker, а ты не забудешь про SchwarzwaldMySQL ?

[schwarzwaldhacker]

Поставить CUDA SDK 3.0/3.1. И закомпилить .

This is a problem. I don't trust 3.0?

This is how I will proceed: I will first release a 7.4 Version, WITHOUT Fermi support(Small speed improvement).

And then I will release the same version compiled with SDK 3.0.

Как всё просто!
Schwarzwaldhacker, а ты не забудешь про SchwarzwaldMySQL ?

As Anna Chapman, my new Russian (Facebook-)Girlfriend would say: "This is irrelevant".

More seriously, let's proceed with one thing. And then proceed with the others. Because I found a couple of smaller errors in SchwarzwaldMYSQL. But unlike SchwarzwaldDES(Bitslice!), I doubt that you can make it much faster.

[schwarzwaldhacker]

I actually finished my next SchwarzwaldDES Version. Could upload it NOW.

Unfortunately, I just found out that my former Webhost got out of Webhosting Business(http://forum.qip.ru/showthread.php?t=40648).

I really need my website in Russia. Of course, because Russia is a great place with beautiful women like Anna Chapman . But also because of legal reasons and other considerations .

If you get me something in North Korea, I might think about it as well. But Russia is really first choice(Outside narod.yandex.ru, please).

... And NO, there are/were no Trojans on my rbcmail.ru website. And there are none in my Crackers.

[schwarzwaldhacker]

OK, just released my latest SchwarzwaldDESBeta:

http://www.schwarzwald.h1.ru/schwarzwaldDES.html

* Better speed!
* More robust, because of less "Pinned Memory", better use of "Shared Memory" and so on.

PLEASE NOTE: I introduced a "Z-FAKTOR", which is basically about the number of inner loops your GPU will do for each launch. Obviously, 1 is best for big lists. But as GPU-Devices are very different, it will be up to you to find the best Z-FAKTORS for your Hashes/Lists.

Удачи!

PS: Sorry if I didn't "greet" everybody. I am kind of very busy. And I just downloaded the new SDK/Drivers. So please, forgive me!

[schwarzwaldhacker]

Please try "bandwidthTest" if you have Fermi!

And please try Xserg's DES on Fermi...

I just successfully installed the new Drivers/Toolkit/SDK. But since I don't have a Fermi, I can't know if I can compile for a Fermi.

Thanks!

[Rolf]

Code: Select all

[bandwidthTest]
BandwidthTest.exe Starting...

Running on...

 Device 0: GeForce GTX 285
 Quick Mode

 Host to Device Bandwidth, 1 Device(s), Paged memory
   Transfer Size (Bytes)        Bandwidth(MB/s)
   33554432                     1586.5

 Device to Host Bandwidth, 1 Device(s), Paged memory
   Transfer Size (Bytes)        Bandwidth(MB/s)
   33554432                     1266.4

 Device to Device Bandwidth, 1 Device(s)
   Transfer Size (Bytes)        Bandwidth(MB/s)
   33554432                     135414.2


[bandwidthTest] - Test results:
PASSED

P.S. Bug report!

Code: Select all

schwarzwaldDES BETA 7.4 REGISTERS Win:v6.01
Written 2010 by Schwarzwaldhacker/ALGORITHM BY XSERG
CudaV1.3  GeForce GTX 285, Processors: 240  Clock rate: 1.548 Mg.
*************************************************
Do you agree NOT TO HARM directly or indirectly
other people with this tool?
(Press y for YES and then return/enter)
y

CharSet mixalpha-numeric-all-space = abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOP
Salt DES(Unix) count 1
CUDA starts with 1 password(s)
:HnkSePmGYydMg:bench              ec Time: 0.01.02
STATUS:8~lch Speed = 1960000 pas/sec Time: 0.01.02
 All checked :-)

MaxAverage Speed = 1327695 Checks per second!(Range/Seconds)
MinAverage Speed = 0 Checks per second!(Range/Seconds)

SchwarzwaldDES found 1 Password(s), 0 Password(s) left

Code: Select all

schwarzwaldDES-registers.exe s=0 e=5 c=mixalpha-numeric-all-space f=hashes.txt o=done.txt b=240 "l=I_read_and_understood_the_licence" z=32

schwarzwaldDES BETA 7.4 REGISTERS Win:v6.01
Written 2010 by Schwarzwaldhacker/ALGORITHM BY XSERG
CudaV1.3  GeForce GTX 285, Processors: 240  Clock rate: 1.548 Mg.
*************************************************
Do you agree NOT TO HARM directly or indirectly
other people with this tool?
(Press y for YES and then return/enter)
y

CharSet mixalpha-numeric-all-space = abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789!@#$%^&*()-_+=~`[]{}|\:;"'<>,.?/
Salt DES(Unix) count 1
CUDA starts with 1 password(s)
:HnkSePmGYydMg:bench              sec Time: 0.00.53
STATUS:$;obh Speed = 17672711 pas/sec Time: 0.00.53
 All checked :-)

MaxAverage Speed = 1553153 Checks per second!(Range/Seconds)
MinAverage Speed = 0 Checks per second!(Range/Seconds)

SchwarzwaldDES found 1 Password(s), 0 Password(s) left

[schwarzwaldhacker]

Rolf, you can use blocks "512". Then for me, loweralpha 1-6 is best with "z=8"(42-43 seconds, GTX+9800). But loweralpha-numeric 1-6 is best with "z=32"(4:59 ).

Also, try this hash: a0000000:bcqgXWKsxYpKM

Because then it will calculate the correct speed for a range, example: (26*26*26*26*26*26)/Time

EDIT: Solved through PM

[schwarzwaldhacker]

And people, PLEASE, try my Fermi samples!

http://3.14.by/forum/viewtopic.php?f=8& ... 8572#p8570

[Rolf]

Well, this is messed up:
:HnkSePmGYydMg:bench sec Time: 0.00.53
STATUS:$;obh Speed = 17672711 pas/sec Time: 0.00.53

[Gajan]

Code: Select all

[bandwidthTest]
bandwidthtest Starting...

Running on...

 Device 0: GeForce GTX 470
 Quick Mode

 Host to Device Bandwidth, 1 Device(s), Paged memory
   Transfer Size (Bytes)        Bandwidth(MB/s)
   33554432                     1019.2

 Device to Host Bandwidth, 1 Device(s), Paged memory
   Transfer Size (Bytes)        Bandwidth(MB/s)
   33554432                     1184.6

 Device to Device Bandwidth, 1 Device(s)
   Transfer Size (Bytes)        Bandwidth(MB/s)
   33554432                     94927.1


[bandwidthTest] - Test results:
PASSED


Press <Enter> to Quit...
-----------------------------------------------------------

Code: Select all

xsergdes b=128 o=output.des.txt l=I_read_and_understood_the_license s=1 e=8 c=dig f=testpasswords.txt
DES(Unix) V2.02 Win:v5.01
1:CudaV2.00  GeForce GTX 470, Processors: 112  Clock rate: 1.250 Mg.
CharSet dig = 0123456789
Restart ---- AT
Salt DES(Unix) count 35
Start with 49 passwords 6-6
X.Z. nVidia driver fall down 0 divice

[schwarzwaldhacker]

Gajan, thanks for your testing. Could you please test Xserg's DES-Cracker as well? I guess:

Code: Select all

XsergDES.exe s=0 e=5 c=loweralpha-numeric f=test.txt

Of course, before you test it, you should create a file "test.txt" with this Hash in it:

Code: Select all

a0000000:bcqgXWKsxYpKM

[Gajan]

Code: Select all

XsergDES.exe s=0 e=5 c=loweralpha-numeric f=salt_SR.txt

DES(Unix) V2.02 Win:v5.01
1:CudaV2.00  GeForce GTX 470, Processors: 112  Clock rate: 1.250 Mg.
CharSet loweralpha-numeric = abcdefghijklmnopqrstuvwxyz0123456789
Restart ---- AT
Salt DES(Unix) count 36
Start with 350 passwords 6-6
X.Z. nVidia driver fall down 0 divice

[schwarzwaldhacker]

Gajan, I am not sure about the results? Could you please restart the test with one password and with "s=1"?

[schwarzwaldhacker]

I will proceed with further tests tomorrow morning(Mon, 19.07.2010). Therefore please help if you have a Fermi!

(We will solve this as well, I am pretty sure )

[Gajan]

unfortunatly same results:

Code: Select all

XsergDES.exe s=1 e=5 c=loweralpha-numeric f=1hash.txt
DES(Unix) V2.02 Win:v5.01
1:CudaV2.00  GeForce GTX 470, Processors: 112  Clock rate: 1.250 Mg.
CharSet loweralpha-numeric = abcdefghijklmnopqrstuvwxyz0123456789
Restart ---- AT
Salt DES(Unix) count 1
Start with 1 passwords 6-6
X.Z. nVidia driver fall down 0 divice

[schwarzwaldhacker]

Gajan, I am sorry about that. There might be 2 problems:

1. I think I compiled it for 2.0(Fermi) AND 1.3. Because if this test had worked, I would only need two files in my releases instead of 3, one for Versions 1.0-1.1, one for Versions 1.2-1.3 and one for Fermi.

2. There might be a Shared Memory problem, particularly with Xserg's DES-Cracker.

Anyway, I will try further tests tomorrow morning. And I would be very happy, if you could help me then!

[Gajan]

Anyway, I will try further tests tomorrow morning. And I would be very happy, if you could help me then!

You got this all wrong my friend!
>I< am the one happy to help
I'm home from work at 16:30 local swedish time.

[schwarzwaldhacker]

In case you are one of the lucky ones who owns a Fermi, please help me to compile my Crackers for Fermi!

I just released a few files and I would be glad if you could test them for me. If things work out, I might then be able to compile my Crackers for your GPU.

http://3.14.by/forum/viewtopic.php?f=8&t=1185

Thanks!

[pwetpwet]

Do you know where i can find a tool like that for Amd Brook (Radeon 5850 HD) please ?

[schwarzwaldhacker]

No, I'm sorry, I don't? Why don't you try Hashcat? It's said to be quite powerful?

[pwetpwet]

it seems to not support DES Crack...

[schwarzwaldhacker]

Ah yes, you are right, didn't know, sorry about that?