Topic: sha1 crack (Read 5581 times)

I'm going to guess this guy isn't interested any more......If you come back I'm still open to offers.

I would also like this, it would provide those of us willing to participate an idea of how much we can earn daily from this. Did you consider my offer that I sent by email?

Hello iulik2k1,

I'd like to test the speed of the SL3 hashing.

Do you have an already solved example that could be used for testing?  For example, what is the input to ighashgpu and what is the expected output?  Or if you don't have a solved example, at least the input (how to run ighashgpu for a specific case of SL3 unlocking)?

The built in test app from ATI is a good start...(I've seen it in Windows, it's probably in Linux too). if I overclock a 5970 to 900 MHz (assuming the system is stable) the test will fill the screen full of speckles and fail.  A speckle is a representation of a stream processor miscalculation.  On the threshold, I will see no speckles, but it will still fail.

Even if your miners are solving blocks, you could still be getting errors and not know it.  Mining is looking for any solution to a problem that has many solutions - errors would merely cost you blocks you never knew you lost, because if you miss one solution, another will come along.  Mining has a tolerance for errors where password cracking does not.  With password cracking, if you miss one solution, it's gone forever and the operation has failed, because there is only one solution (and you'll never know if you missed it either).

I would simply recommend the clock speed that consistently passes the overclock stress test offered by AMD/ATI.  If it fails even once out of twenty, don't password crack at that clock speed.

If the buyer wishes I can soon as go back to clock speeds, but my miners solved a block after 10 hours yesterday so pretty certain they are running okay.

Are there any specific apps which can stress two 5970's at 100% and determine their stability?

Exactly, but I wouldn't call it artificial.

I would assume this means that it would artificially depress the mining difficulty while having no effect on the exchange rate, as people leave the mining pool to sell their "supercomputing" services on the open market to an influx of password crackers (or other buyers of computing time).

This side effect of Bitcoin is pretty novel - it has gotten us - a whole bunch of average joe people not otherwise hobbyists of supercomputers - to become interested in distributed computing even if we don't realize it - and to get us set up in a way where we can quickly and easily rent teraflops of computing power to random people without a second thought.  We've even got an anonymous and foolproof way for them to pay us for it.

Without us, people otherwise have to have access to expensive private computing resources or botnets (or buy lots of 5970s themselves) if they want to run huge computational tasks, and for most people, if it's just to solve some short term problem (like a password crack) it isn't worth their while.  We represent a new low-cost option.

Just curious, do your overclocked 5970's pass all diagnostic tests without display speckling at 900 MHz?  When you overclock a GPU it may appear to function but it may be malfunctioning and you have no way to know.

When you overclock a CPU, the consequences are that it will freeze if it's too fast - it can't continue to function if it's making processing mistakes.  When you overclock a GPU, it may not freeze, it just may make mathematical errors you may never detect because most of the output is being discarded if it doesn't satisfy some criteria.  In the context of gaming, you get specks on the screen or other artifacts.  In the context of bitcoin mining, you might be getting 350 MHash/sec on the display, but a certain percent of those may be miscalculated and worthless.

On Bitcoin, it's no real loss to have a few glitched calculations because it's the total number that matters.  i.e. as long as you're gaining more good hashes by overclocking than you're losing by doing so, you're basically ahead.  For every block, there's a huge number of solutions that will solve it, you only have to find one.

But if you're password cracking, you need 100% accuracy, because there is only ONE solution to the problem, and you can't afford to miss it.  If GPU "glitches" a calculation that happens to belong to the right password, it will falsely report the password wasn't found when it was.  That undermines the total distributed cracking effort.

Also, a password cracker is likely to use more resources on the GPU especially if doing a dictionary attack (specifically video memory - bitcoin mining uses practically none).  The need for these extra resources may tip over a GPU that otherwise mines bitcoins just fine.

Care to ellaborate? 

This type of enterprise can negate the association between difficulty and exchange rate, if there ever was one.

Will run SL3 code at $25 per 24h. I can start straight away but I want a confirmed payment in advance before I begin (How do I know your going to pay at all). I will want each $25 in advance, feel free to send a bulk payment for consecutive days. I don't want to do a "Pay-per-code" deal at this time as I need a more stable income. I currently earn around $23/24h from mining so the $25 is an incentive for me to switch.

Got a deal? If so I will provide my bitcoin address and as soon as payment is received I will begin processes what you like with IGASHGPU.

Just to confirm I have two 5970's overclocked to 900Mhz so my hashing rate is 20% higher than normal.

hi, we calculate sl3 nokia code with igashgpu, the curent price for one code is 15-18$ to 22e
igashgpu working with 5970, 58xx, 57xx, not working with 6xxx series ati card
on nvidia card is working very slow.

igashgpu is best for sha1 bruteforce. is also a new version of igashgpu with cut time to half but the version is not available to public.

how many bitcoins you do on 24h? can you sell for what price?
if the price is good for sl3 then do sl3 unlock code.

http://golubev.com/gpuest.htm


is interesting to do client-server solution with client can do sha1 or sha256 btc.
the client can pay with money or btc.
this can add more value for btc.

If his problem is just "please find me the string with this SHA1 hash" then you could probably just find and configure the SHA1 cracker of your choice (avoiding security concerns) and make him an offer.

He would also need to specify what key space he thinks the password would be, and why he thinks the password should be within that key space.  He mentions 4 GPU-days of 5970, but that doesn't exhaust all of SHA1 (160 bits), that merely exhausts a certain key space which may or may not contain the password he's looking for.

One cracker for Windows is called IGHASHGPU and claims to deliver 2300 Mhash/s (SHA1) on a 5970.

I have two 5970's available, and would much rather have a fixed income then than variable mining at this moment in time.

My cost of electricity per day is around £4.08. Make me an offer and I will consider it.

Also is your password cracking program run under windows or linux? If it's linux I can install it, but you will need to make a guide to configure all the settings.
 

I can imagine mining pools getting into this business.

let me know your hash and I will look for it in my rainbow tables.

You could probably find people here with 5970's who would rent you a fixed amount of computing time on their 5970's for BTC if you were to provide whatever was supposed to be run on them.

hi, to all bitcoin users.

i'm landing here searching for 4xhd 5970.
i need gpu power machine to crack passwords.
i  can pay for code finded.
for 1 ati hd5970 it's take 4 day for 1 code.
with 4 ati will take 1 day.
how much money do you want to calculate code for me?

how many bitcoins you win on one day? how much is worth?
if it worth to make bitcoins i will buy a new computer.






my email [email protected]