Search found 111 matches

by D3ad0ne
Thu Mar 31, 2011 12:04 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: MD5 45.1 billion hash/sec
Replies: 5
Views: 15070

Re: MD5 45.1 billion hash/sec

Thanks BarsMonster. I hadn't actually heard about this unlocking of 6950.
If I had it might have saved me some on buying these cards :) Do you have a link?
by D3ad0ne
Mon Mar 28, 2011 1:44 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: MD5 45.1 billion hash/sec
Replies: 5
Views: 15070

MD5 45.1 billion hash/sec

Just got the new machine built, with some new speeds on MD5, NTLM, MD4, and SHA1 MD5 45B/sec NTLM 70B/sec MD4 75B/sec SHA1 14B/sec Here is the vid - http://ob-security.info/files/oclhc-lite.avi <-- need divx codec This is using beta software oclHashcat-lite which will be released soon. System is a T...
by D3ad0ne
Fri Dec 10, 2010 5:13 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: 33.1 billion hash/sec
Replies: 25
Views: 37670

Re: 28.6 billion hash/sec

I been able to use a x1 slot with an x16 card before. I haven't tried your method though.

There has been some talk about your application. If you get a chance you should join us in the #hashcat channel on irc.rizon.net
by D3ad0ne
Thu Dec 09, 2010 6:03 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: 33.1 billion hash/sec
Replies: 25
Views: 37670

Re: 28.6 billion hash/sec

Nice man, I love a challenge. To bad I can't run this tool on my system.

Also after reading your site you have stats from a 8 GTX580 system. Who has that?
by D3ad0ne
Tue Nov 30, 2010 2:38 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: BarsWF gone opensource
Replies: 14
Views: 76303

Re: BarsWF gone opensource

:shock: Wow.. Svarychevski let me be the first to say thanks. BarsWF was one of the first programs that got me into GPU cracking. Hopefully you continue to enjoy whatever it is your doing, and best of luck. :D
by D3ad0ne
Fri Nov 12, 2010 3:28 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Qestion
Replies: 9
Views: 14918

Re: Qestion

Just because it can be done, doesn't mean it's the smart way to do it. Would you want to crack 5000 md5 hashes one at a time or all at once.. If the keyspace you were searching took on average 1 min to finish per hash it would take you 3.4 days to finish 5000 hashes in single hash mode. Even if it t...
by D3ad0ne
Wed Oct 27, 2010 10:16 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Challenge: Can you beat these hash/sec speeds?
Replies: 5
Views: 15601

Re: Challenge: Can you beat these hash/sec speeds?

Well I see no one has arisen to the challenge. Not to make it any easier I raise the bar a bit higher. Here is the final addition to my current build.

8 GTX 480's running at 1711 shader speeds.
IGhashgpu = 17239.7M
BarsWF64 = 14071.6M
ig17B.jpg
ighashgpu v.90
Bars14B.jpg
Bars14B.jpg (110.2 KiB) Viewed 15273 times
by D3ad0ne
Mon Oct 25, 2010 3:37 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: FRT CUDA client
Replies: 3
Views: 9289

Re: FRT CUDA client

Cool, any word on when they'll have a GPU version of of rcrack?
by D3ad0ne
Sat Oct 16, 2010 1:19 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Ighashgpu .90 Beta new speeds
Replies: 0
Views: 7807

Ighashgpu .90 Beta new speeds

IvanG released a new beta version of ighashgpu. He talks about it in his blog - http://www.golubev.com/blog/?p=140. From what I understand he added some code for Nvidia that supports the SHL+ADD. I tried it out and it was a 3% improvement for me but every little bit helps! Then I went a little crazy...
by D3ad0ne
Mon Oct 04, 2010 4:29 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Hit the 8 Billion hashes /sec mark
Replies: 14
Views: 23605

Re: Hit the 8 Billion hashes /sec mark

Sorry for the delay in reply. Blazer: It runs off of two 1200watt power supplies. And it will play Crysis with quad sli, however it is a bit of a pain to play games because I have to disable the extra GPU's thankfully the motherboard has jumpers that do this without requiring pulling out the cards, ...
by D3ad0ne
Wed Sep 15, 2010 4:20 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Challenge: Can you beat these hash/sec speeds?
Replies: 5
Views: 15601

Challenge: Can you beat these hash/sec speeds?

Alright finally got all the cards working in one system. Just wanted to challenge anyone to see if they have a faster system.
BarsWF64.jpg
BarsWF64.jpg (46.44 KiB) Viewed 15601 times
ighash_12B.JPG
by D3ad0ne
Sat Sep 11, 2010 5:42 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Hit the 8 Billion hashes /sec mark
Replies: 14
Views: 23605

Re: Hit the 8 Billion hashes /sec mark

If I can get all of these to work together I'll see what we can get :)
cards.jpg
6 GTX 480 hydrocopper's
by D3ad0ne
Sun Sep 05, 2010 5:07 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Hit the 8 Billion hashes /sec mark
Replies: 14
Views: 23605

Re: Hit the 8 Billion hashes /sec mark

Would 4x 5970 beat that?
Good question, anyone have a 5970 that can show their GPU speed?
by D3ad0ne
Sat Sep 04, 2010 10:52 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Hit the 8 Billion hashes /sec mark
Replies: 14
Views: 23605

Hit the 8 Billion hashes /sec mark

Well I've been working on the new machine and ran BarsWF 64 to get some bench marks. This is a SR-2 Classified MB with two xeon x5650's. With 4 hydro copper GTX 480's. I had to over clock the GPU's a bit to hit the 8B mark but even running at 1711Mhz shader speed everything was running below 40c tha...
by D3ad0ne
Mon Mar 29, 2010 9:05 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Nvidia GTX 480 to only have 480 cores.
Replies: 2
Views: 3483

Nvidia GTX 480 to only have 480 cores.

As some of you may know the GTX 480 and 470s have been released. Upon inspection of the stat sheet it turns out that the 480 will not have the 512 cores it was orginally thought to have instead it will only have 480 cores enabled running at 1401 Mhz. gpu_cores.JPG For anyone that is curious that sho...
by D3ad0ne
Fri Jan 15, 2010 3:42 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: speed of BarsWF / other
Replies: 12
Views: 8530

Re: speed of BarsWF / other

I did some testing again recently of several of the top MD5 hash crackers and BarsWF pretty much holds the lead. Mainly because it is using both the CPUs and the GPUs. IGhasheGPU v0.70 seems to be faster compared to BarsWF when comparing GPU only in both CUDA and ATi. IGhashGPU v0.70tv = 408M cputdo...
by D3ad0ne
Thu Jan 14, 2010 9:07 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: CUDA supported bruteforcers
Replies: 35
Views: 24203

Re: CUDA supported bruteforcers

vampyr, I'll do some independant testing later this week.
by D3ad0ne
Tue Jan 12, 2010 4:13 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: CUDA supported bruteforcers
Replies: 35
Views: 24203

Re: CUDA supported bruteforcers

i'm looking to build a reasonable sized network specifically for password analysis research Could you be a bit more specific about the research your looking to do? I agree with Ivan, it really dpends on what types of passwords or research your wanting to do. If you ask me if you want a universal se...
by D3ad0ne
Wed Dec 30, 2009 12:34 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: GSM - A51 Rainbow Table for mobile decryption
Replies: 34
Views: 25324

Re: GSM - A51 Rainbow Table for mobile decryption

Neat, this was from today. The article says: he had deciphered and published the secret code used to encrypt most of the world’s digital mobile phone calls about 24 people, some members of the Chaos Computer Club, which is based in Berlin, worked independently to generate the necessary volume of ran...
by D3ad0ne
Sun Dec 27, 2009 10:23 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: GSM - A51 Rainbow Table for mobile decryption
Replies: 34
Views: 25324

Re: GSM - A51 Rainbow Table for mobile decryption

Ok well apparently it doesn't use all the GPUs by default. So I think I got all the GPUs working together now: Initialize implementation sharedmem... Initialize implementation sharedmem... 0 chains done, current rate 0.00 chains/sec (interval: 00:01:00) 4602 chains done, current rate 76.70 chains/se...
by D3ad0ne
Sun Dec 27, 2009 8:47 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: GSM - A51 Rainbow Table for mobile decryption
Replies: 34
Views: 25324

Re: GSM - A51 Rainbow Table for mobile decryption

We have the tools which are opensource and the scanner is a USRP (u$s 700.-) What are some examples of scanners/software that can be used to do this? Links, model #s, etc? I'm getting rid of my pro-164 and am looking to get something a little more versatile. I thought I would check this out.. So i ...
by D3ad0ne
Sun Dec 27, 2009 1:56 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: hashcat
Replies: 7
Views: 9559

Re: hashcat

For those that haven't checked this out yet, this program is like John the ripper and passwordspro combined but so much more! I have been using it for the past couple months and once you get a feel for some of the options and how it works, the versatility of it is simply amazing! Granted it is not G...
by D3ad0ne
Mon Dec 21, 2009 7:15 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: GPU temperature
Replies: 7
Views: 6445

Re: GPU temperature

you may start with replacing thermal grace under cooler Good suggestion, I have replaced all the thermal compound on my GPUs and CPUs. I used to get above 95ºc with stock compound and now it is well below 90ºc at full load. I recommend diamond compound because it has a much higher thermal conductiv...
by D3ad0ne
Wed Dec 16, 2009 2:47 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: 9+ GPU machine
Replies: 12
Views: 11615

Re: 9+ GPU machine

Thanks for the link Ivan. It's funny that they point out: Although the system is up and running, we are still experiencing software stability issues, probably caused by an incompatibility between the video drivers and the BIOS and Linux modifications we had to use So I'm glad to see I'm not the only...
by D3ad0ne
Tue Dec 15, 2009 9:00 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Maths behind MD5
Replies: 12
Views: 8394

Re: Maths behind MD5

In Diffie-Hellman if p is not prime bad things happen because b needs to be a primitive root mod p. Which mean that for any n: m = b ^ n (mod p), m is co-prime to p. Now if p is prime then m can be any number from 1 to p-1 but if p is not prime then there are "holes" in that range that m can not be...