EvilZone
Hacking and Security => Hacking and Security => : eliaou December 01, 2013, 11:32:48 PM
-
Hi Guys ,
I started to get interested in Cryptocurrency mining and the basic concept ( I guess most of you allready know ) is that you find a solution to a hash a bit like brute forcing , now miners figured out that actually GPU's (and later technology witch are asic chips that run at +- 500 gigahash /s instead of CPU's witch work at max 1 megahash/s ) are alot faster at solving this kind of problems than CPU's are .
My question is : would it be faster to brute force a hash with CPU or GPU for anything ?
link (simple diagram about Cryptocurrency mining) : [size=78%]http://imgur.com/IdlzmLA (http://imgur.com/IdlzmLA)[/size]
Thank You
-
GPU faster than CPU.
I mean, you answered your own question...
-
First of there are multiple flaws in your post.
GPU mining is NOT a new thing, GPU mining started in 2010 or 2011 afaik. Then FPGA chips was used, and now a bit more recently (but still nothing new) ASIC chips are used. Also 500 GH/s +- is not very accurate either. 500 GH/s is pretty good atm. And finally CPU's are not limited to one MH/s, mine does about 6 MH/s and I am sure one can do a bit better than that as well.
I am not sure I understand your question. CPU < GPU / FPGA < ASIC. GPU will pretty much always be faster than CPU.
-
CPU mining is absolutely worthless at the difficulty rate that Bitcoin is now at. This is mining 101.
On the other hand, it's still feasible for cryptocurrencies like Litecoin, which uses a different proof-of-work scheme.
-
What i meant is using it to brute force any regular password lets say
-
What i meant is using it to brute force any regular password lets say
Using Bitcoin ASIC's to brute force other algorithms? I don't think so, not without heavy modification. The whole point with ASIC's is that they are application specific and made for that application only. This is what makes them so fast. But obviously you could make a ASIC for other algorithms but I don't know of any on the top of my head.
If you are still wondering: GPUs are much faster than CPUs for general password cracking as well.
-
ok thanks..would it be just modifications in the driver ?
-
ok thanks..would it be just modifications in the driver ?
I honestly don't know for sure. But I would guess this depends on the chip and the extent of the desired modification. ASIC's are created fully or partially custom for one specific task. So I would imagine if you wanted to make a MD5 bruteforcer, a fully custom ASIC for the SHA256 algorithm would be useless without hardware and software modification, which would be substantial modifications as MD5 and SHA256 doesn't have a lot in common.
-
ok..so no brute forcing for me I guess :'( .
Thanks for the help anyways
-
ok..so no brute forcing for me I guess :'( .
Thanks for the help anyways
If you're interested in brute forcing, you should look at the possibility of running multiple GPUs on the same machine (can't remember the technical term). Also, distributed hash cracking is very interesting.
-
Yes I tought of parallel computing ,but (sorry for the noobnes here ) is it possible to divide a hash by segment like so : core 1 does "2j" core 2 does "h2".. etc
Or just lets say : core 1 does lowercase core 2 : uppercase ...etc
-
yea I get it parallel computing, but (sorry for the noobnes here) is it possible to divide a hash by segment like so : core 1 does "2j" core 2 does "h2".. etc
Or just lets say : core 1 does lowercase core 2 : uppercase ...etc
You cant divide the hash but you can perform different types/ranges of brute force on the same hash on different cores.
-
Its a really interesting subject.
But if peoples do it successfully ,can lets say sha-256 be still secure ? because its still
(2^64 ) -1 possible combinations and with a few asic chips and doing some parallel hashing peoples could brute force hashes pretty quick no ?
-
Its a really interesting subject.
But if peoples do it successfully ,can lets say sha-256 be still secure ? because its still
(2^64 ) -1 possible combinations and with a few asic chips and doing some parallel hashing peoples could brute force hashes pretty quick no ?
I have yet to see a SHA256 ASIC chip. But considering the best ones (Bitcoin) does about 500GH/s now, that's 500 000 000 000 attempts per second. A 10 letter password with upper and lowercase letters would still take ~349774 seconds or ~5829 minutes or ~97 hours or ~4 days. And that's with at least 500GH/s 24/7. Add one number to that password and you have 22 days.
Personally I don't use SHA or any variation of SHA alone anymore. I do multiple rounds with a feedback hash from the previous round and salt.
-
A hash algorithm's security isn't measured so much in how fast you can bruteforce it (that's more of a side effect), but rather flaws in the actual digest: collisions (primarily), preimage attacks and length extensions (in the event that the hash is based on a Merkle-Damgard construction, most are).
All of these flaws then contribute to how easily and quickly it can be bruteforced, as they show poor entropy.
-
Entropy ? (again sorry for the noobness )
Do you mean that by definition a password becomes more complex and has many layers of change til it becomes a hash ? and by collision do you mean that there can be a recognizable pattern how the entropy develops , and by that the computer arrives at a conclusion as to what the original password or the prime number was ?
(by the way i love this forum lots of smart peoples here , its refreshing to see that it still exists in the internet :) )
-
Entropy - The measure of randomness in a hash (In this case anyway)(Someone correct me if i'm wrong).
Basically you want a hash function to have a completely different output even if the input is only changed a tiny tiny tiny bit. So that hash of aaaaa is not even CLOSE or similar of hash of aaaab.
Collisions is when you have two different inputs that produce the same output in a hash function. There is no such thing as a hash without a collision, this is because a hash is of fixed length. However, the entropy of a hash determines "how often" a collision will occur or "how good" an hashing algorithm is.
(Someone please correct me if I am mistaken above, its getting rather late over here)
-
Thanks for the info ande .
I got alot more questions but I think I can continue this one on google
+1 cookie
-
One of the main reasons CPU mining is slow (rather than saying gpu mining is fast) is because CPU implements a lot of security features and has more features who come in handy in running an operation system and securing it. The GPU is more based on raw calculations wich makes it faster.
-
Security in the CPU itself ?
Do you mean that it has security in the gate transistors or/and in the logic circuit ?
-
Security in the CPU itself ?
Do you mean that it has security in the gate transistors or/and in the logic circuit ?
Yhea a CPU has rings and modes http://en.wikipedia.org/wiki/Ring_(computer_security). Also a CPU has virtual memory and more features wich are required to run a OS but not required to do raw operations.
-
Entropy - The measure of randomness in a hash (In this case anyway)(Someone correct me if i'm wrong).
Basically you want a hash function to have a completely different output even if the input is only changed a tiny tiny tiny bit. So that hash of aaaaa is not even CLOSE or similar of hash of aaaab.
Collisions is when you have two different inputs that produce the same output in a hash function. There is no such thing as a hash without a collision, this is because a hash is of fixed length. However, the entropy of a hash determines "how often" a collision will occur or "how good" an hashing algorithm is.
(Someone please correct me if I am mistaken above, its getting rather late over here)
Nah you're not wrong. This is part of the reason why an algo is only so good for so long. People used to think it would take a Cray supercomputer years to crack DES.