[oclHashcat-lite] WPA/WPA2
#1
Hello, I think, it's good idea to implement WPA/WPA2 into oclHashcat-lite. Since oclHashcat-lite is focused at one-shot cracks and it supports up to 54 characters long passwords, it'd be good to implement this algo., when even descrypt is implemented. I have speed of 15500 k c/s on WPA with oclHashcat-plus on my ATI Radeon HD 4850 and on oclHashcat-lites site we can see, that descrypt has speed of 19400 k c/s on NVidia gtx560Ti, which is as fast as my graphics card.
What do you think ?
#2
this limit has nothing to do with plus or lite. Implementing in lite would only mean you are limited to mask attacks.
#3
So why is descrypt implemented in oclHashcat-lite ?
#4
what does descrypt have to do with wpa?
#5
descrypt is as slow as WPA or even slower
#6
i think you miss the point of lite. the wpa kernel in plus is already optimized for single hash. it would not be any faster in lite. all you would get in lite is no wordlist support, which is practically worthless.
#7
(02-01-2013, 10:43 PM)Kuci Wrote: Hello, I think, it's good idea to implement WPA/WPA2 into oclHashcat-lite. Since oclHashcat-lite is focused at one-shot cracks and it supports up to 54 characters long passwords, it'd be good to implement this algo., when even descrypt is implemented. I have speed of 15500 k c/s on WPA with oclHashcat-plus on my ATI Radeon HD 4850 and on oclHashcat-lites site we can see, that descrypt has speed of 19400 k c/s on NVidia gtx560Ti, which is as fast as my graphics card.
What do you think ?

I share your enthuasium for increased speed or password length for WPA.

However, as undeath mentioned, you would be limited to mask attacks in lite.

Unless you are using Markov with an extremly good hcstat file, combined with a lot of luck I doubt you would achieve much success trying to brute 15 + characters in WPA.

I am not trying to put this idea down but I just don't see much of an advantage without word list support then we are back to oclhashcat really. There are a few people who are looking forward to greater password lengths in oclhashcat and I think we have managed to persuade atom to consider it.

As for a speed increase below 15 characters in lite, I doubt there would be much increase but I am no expert. I think atom has squeezed every bit of optimisation out of oclhashcat and I think we must have reached the end of software performance increases with WPA.

I think the hard fact is this, if you want faster WPA cracking it is probably time to buy a better GPU.

I feel your pain, I really do, WPA is my favorite algo.
#8
Sorry, I didn't notice epixoip's post just before mine, I was typing. I have pretty much repeated what he has said and now I look retarded.... even more than I normally do. Smile
#9
Oh guys, I forgot, that in plus it's already optimised for one hash only. You've got the point.

@Hash-IT: mate, you're all right, just look at me Big Grin Programming for 5 years in several languages and still do that stupid mistake which forgetting is Big Grin