Search found 82 matches

by neinbrucke
Mon Jan 12, 2009 7:24 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: about the algorithm used
Replies: 49
Views: 39355

Re: about the algorithm used

i use part of the rainbow tables reduction function to calculate the starting points for each keyspace.
by neinbrucke
Sun Jan 11, 2009 9:19 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: about the algorithm used
Replies: 49
Views: 39355

Re: about the algorithm used

oh lol... and probably you are generating strings as input... not even an array of characters... strings are slooooow. i'm doing sse2 in vc++ 2008, i should give the intel compiler a try :P with my sha1 brute forcer i'm now from 5M hashes/s without SSE2 to 10M hashes/s with SSE2... should still be a...
by neinbrucke
Sun Jan 11, 2009 8:30 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: about the algorithm used
Replies: 49
Views: 39355

Re: about the algorithm used

ah ok, that would be for a reversed md5 i think?

i'm doing sha1... doesn't seem to reverse that well :) (data is being 'expanded' after step 16)

just curious, did you start on an optimized sha1 cracker already?
by neinbrucke
Sun Jan 11, 2009 12:58 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: about the algorithm used
Replies: 49
Views: 39355

Re: about the algorithm used

oh, nice solution :)
i doubt it will be faster to do a PTEST actually...

btw wouldn't it be faster to only compare as1 and tas1? and only on success start comparing the other variables?
by neinbrucke
Sat Jan 10, 2009 9:54 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: about the algorithm used
Replies: 49
Views: 39355

Re: about the algorithm used

i've been toying around with sha1 and sse2 (first time i use sse :P), and i'm sort of learning it (already do 4 hashes at the same time, not optimized at all, but still like 50% faster overall)... do you use sse2 for generating plaintexts as well? also, is it me, or is there no PTEST like function i...
by neinbrucke
Fri Jan 09, 2009 10:14 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: about the algorithm used
Replies: 49
Views: 39355

Re: about the algorithm used

not sure where your slow down is, probably at multiple places... my sha-1 brute forcer already does 5M hashes/s (well, that is a 3.2 ghz core, but still), while sha-1 is slower then md5... and it seems like you can't reverse as much as with MD5. oh btw, if i just use openssl's MD5, i already get mor...
by neinbrucke
Fri Jan 09, 2009 9:20 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: about the algorithm used
Replies: 49
Views: 39355

Re: about the algorithm used

your program probably doesn't have SSE2 optimizations... and you probably don't reverse your hashes to a certain level :) brunolap, how fast is your bruteforcer when you check only one hash at the same time? does it scale well? the comparison part can also be a big hold up. but i am somewhat curious...
by neinbrucke
Mon Jan 05, 2009 8:54 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: What hash to support next?
Replies: 51
Views: 36073

Re: What hash to support next?

tronador wrote:NTML :P

I manage some databases with passwords md5(pass), i think i'll change to md5(sha1(pass)) :lol:
if you are going to change it, how about using sha2-256(salt.pass) or something like that :)
by neinbrucke
Sat Jan 03, 2009 8:01 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: road map.
Replies: 20
Views: 14310

Re: road map.

hmm, true about distributed projects :) so you'll probably be doing statistics and teaming as well? :P i do think it will be very hard to find a good solution for hashes from legal penetration tests. i think for your supercomputer, it is better to not try and handle that aspect, as it probably compl...
by neinbrucke
Sat Jan 03, 2009 4:42 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: road map.
Replies: 20
Views: 14310

Re: road map.

tnx for your answer. my main concern is number 1 and 3 (and in several cases 4 and 5, although they strongly relate to 1 and 3). Whether it would be hashes from a legal penetration test or from some personal website, there is no way to make sure who has access to these hashes (and recovered password...
by neinbrucke
Sat Jan 03, 2009 3:12 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: road map.
Replies: 20
Views: 14310

Re: road map.

As I was saying several times, stand-alone version with multi-hash support will never be released. It would be supported only in distributed version. You can see why I am doing that. i was already afraid that was what you meant... i will probably never use a (public) distributed cracker for any has...
by neinbrucke
Tue Dec 30, 2008 11:15 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: What hash to support next?
Replies: 51
Views: 36073

Re: What hash to support next?

i think next to bsd md5, crypt des would also be nice. wpa would be awesome, but i guess it would hold up the full distributed network and maybe just get loweralpha with length 8. Or there should be some sort of wordlist option, probably making your implementation completely different (and maybe not...
by neinbrucke
Tue Dec 30, 2008 11:09 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Charsets to work on in distributed version
Replies: 15
Views: 12929

Re: Charsets to work on in distributed version

at least don't use the lm-frt charsets... they are meant for LM hashes, where unicode characters are first mapped to certain 'ascii' characters (actually just single byte chars). The mapping differs per OEM code page, so US english has another mapping then Latin (most western europian countries). Th...
by neinbrucke
Sat Dec 27, 2008 7:45 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Status of the ATI GPU Version ?
Replies: 224
Views: 175511

Re: Status of the ATI GPU Version ?

strange indeed :)
but this case is not on it's own... VC++ has more weird differences between debug and release... (just can't recall what my problem with it was some months ago :P)
(and tnx for the explanation)
by neinbrucke
Sat Dec 27, 2008 1:20 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Status of the ATI GPU Version ?
Replies: 224
Views: 175511

Re: Status of the ATI GPU Version ?

no, i like the talk :P

were the problems introduced when you started using boost? or is that just coincidence?
do i understand right, result of floorl is first cast to an int64_t, then to int.... why the step in between ?
(btw, my c(++) isn't that great, so i may be asking stupid questions ;))
by neinbrucke
Fri Dec 26, 2008 9:55 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Status of the ATI GPU Version ?
Replies: 224
Views: 175511

Re: Status of the ATI GPU Version ?

yeah, i use it for a multithreaded version of rcracki. i first used windows threads, then thought it would be nice to have multiplatform support... it seemed to me that boost is more 'high level' then pthreads (and windows threads). but i could be wrong, i just choose pthreads as it seemed like a fa...
by neinbrucke
Fri Dec 26, 2008 11:49 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Status of the ATI GPU Version ?
Replies: 224
Views: 175511

Re: Status of the ATI GPU Version ?

is boost that much easier or more powerful then just using posix threads? :)
by neinbrucke
Sun Dec 21, 2008 12:57 am
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: BarsWF feature wishlist
Replies: 53
Views: 45183

Re: BarsWF feature wishlist

wordlist mutations are VERY powerful... so i think it is a valuable item for the wishlist :)
by neinbrucke
Wed Nov 26, 2008 10:32 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Status of the ATI GPU Version ?
Replies: 224
Views: 175511

Re: Status of the ATI GPU Version ?

you've got some really anxious (and a little mental) fans bars :P
by neinbrucke
Wed Nov 19, 2008 10:33 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: -c 0a command
Replies: 5
Views: 5449

Re: -c 0a command

it should find both, alpha-numeric also includes the passwords with just numeric or just alpha
by neinbrucke
Sat Nov 15, 2008 7:19 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Status of the ATI GPU Version ?
Replies: 224
Views: 175511

Re: Status of the ATI GPU Version ?

there is not yet a working ATI GPU version... and besides, you seem to be running the SSE2 version, which uses only CPU anyway
by neinbrucke
Thu Nov 13, 2008 3:28 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Poll: How many hashes you are going to crack in multi-hash?
Replies: 7
Views: 8319

Re: Poll: How many hashes you are going to crack in multi-hash?

quite often i try hundreds or thousands of unsalted md5 hashes at once... cuda_nv worked fine with around 20k hashes.
but i wouldn't mind being able to crack > 100k hashes at once with barswf :)
by neinbrucke
Thu Nov 13, 2008 3:26 pm
Forum: General Talk
Topic: Quadro FX 5800
Replies: 8
Views: 9308

Re: Quadro FX 5800

would that really be useful for rainbowtables? you only need to store endpoints, but i don't know how often you can transfer your generated chains from GPU memory to the host without creating too much overhead. (as i understand, those copy actions are about the most expensive actions on doing GPU st...
by neinbrucke
Mon Nov 10, 2008 9:16 pm
Forum: BarsWF, Cryptography, Security, GPGPU and supercomputing
Topic: Request: NT Hash (actually MD4)
Replies: 54
Views: 59221

Re: Request: NT Hash (actually MD4)

However, 20M * 1000 = 20B, which is my effective check rate. oh ok, than i understand you... but it's kind of strange giving that number as the effective rate... as this counts for all brute forces on unsalted hashes... in my opinion the actual strength of a brute forcer is in it's speed of generat...