Topic: BitCrack - A tool for brute-forcing private keys - page 34. (Read 74707 times)

I forked now BitCrack and improved it a little.

https://github.com/Uzlopak/BitCrackOpenCL/

In this Fork I concentrate on the OpenCL implementation. So no CUDA-stuff.

I already could improve the performance about 30 %. I assume I can get it better, when I start vectorizing the operations.

Already wrote on stackoverflow for help to get atleast the multiply256 function improved:

https://stackoverflow.com/questions/67667314/transform-native-c-matrix-multiplication-to-opencl-simd-matrix-multiplication

Already posted in various chats and on fiverr with a bounty of 25 € to get it done by somebody else... Probably will ending solving it by myself. Some dude on fiverr was first promising to implement it, but I guess it was more like a dud.

I assume a performance improvement of 200-300% when this function is using SIMD-Operation.

I am fixing and optimizing the OpenCL Version of the Bitcracker... It is really odd, but the debug version is faster than the release version... Crazy.

Any Ideas?

Yeah ! faster .. see your screenshot - https://ibb.co/Dpc0HxR -  50% in 1.26609e+32years ^^ good luck  so DON'T BUY IT !!

It's VanitySearch...
But you need to understand the BIG differences between this 2 programs :

VanitySearch : Search patern of bitcoin address with random search.
It's use the symetrie of the elliptic curve. Ex :
privKey "1" -> Calc puKey = X, Y -> Sha256 -> Ripemd160 -> found ? no : puKey for "-1" = X, -Y -> Sha256 -> Ripemd160 -> found ?
privKey "2" -> Calc puKey = X, Y -> Sha256 -> Ripemd160 -> found ? no : puKey for "-2" = X, -Y -> Sha256 -> Ripemd160 -> found ?
...
So the speed is highter than Bitcrack because for 1 pubKey calc you have 2 result.
But if you search in a specific range like "BitcoinChallenge", 50% of the search no use...

BitCrack : Specific bitcoin address in specific range with ordered search.
It don't use the symetrie of the elliptic curve. Ex :
privKey "1" -> Calc puKey = X, Y -> Sha256 -> Ripemd160 -> it's equal ?
privKey "2" -> Calc puKey = X, Y -> Sha256 -> Ripemd160 -> it's equal ?
...
So the speed is hight for specifique range. But you need to know your graphic card to configure the program.
For exemple, with a gtx 1066, the SM count is 10 so the param should be : -b 10 -t 256 -p 1024 or -b 20 -t 256 -p 768 or -b 100 -t 256 -p 512

That program is, or is using, VanitySearch Program created by Jean Luc Pons AND Bitcrack. First program is VanitySearch and second program is Bitcrack. Nothing new there.

https://github.com/JeanLucPons/VanitySearch

What is this program? It generates random addresses at a speed of 240m / s on gtx1066 and checks against the base of addresses, while for bitcrack only 25-30 in the -u -c mode and only with the specified range. I found a video on YouTube and took screenshots.

screen link
https://ibb.co/qpz15HG
https://ibb.co/9wzzVDZ
https://ibb.co/Dpc0HxR
https://ibb.co/YWQcqtM
https://ibb.co/rZ1XPT8

That is because the majority of the transactions before #64 were solved by this tool https://bitcointalksearch.org/topic/large-bitcoin-collider-thread-20-1877935 long before Bitcrack was invented.

Yeah obviously you don't really understand the puzzle or math...100 people with one GPU, or two GPUs or maybe three GPUs??  There has been a pool set up for years, where I know for a fact over 300 GPUs have worked on #64 for weeks at a time.  All combined, over a few years, only around 15% of the range has been searched.

I have not found a single comment on a successful search, neither in this thread nor in the main one! It looks like this is a puzzle of cheating and trolling over people, otherwise there would be at least a few successful treasure hunters in the early ranges up to 64.
It looks like the range is incorrectly specified and can lie anywhere except the indicated one. 64 for a long time could be grouped into 100 people and sort out completely in a week or two.

You can install C++ Clang tools for Windows on the community addition also

So the Pro edition bundles a clang compiler? That could actually be useful for people who want to avoid screwing around with MSbuild since you can make both Windows and Linux binaries from a Linux system (and vice versa).

NVIDIA GeForce RTX 2080

I think driver problem
first try update CUDA driver to 10.2

other try compare old release
https://github.com/brichard19/BitCrack/releases/tag/0.31
https://github.com/brichard19/BitCrack/releases/tag/0.30
https://github.com/brichard19/BitCrack/releases/tag/0.29

Which driver version are you using? You will probably have to roll back your driver to an earlier version. That is how I got it to work months ago.

Well now I bought VS 2019 Pro with small money and used CLANG for compiling clBitCrack. Got from 190 MKeys/s to 192 MKeys/s.

I think this says all.

GPU is - GTX STRIX2080OC
CUDA is 10.1
I followed all the instructions I can find. It does the same thing even with just using an address.
 https://ibb.co/YjmTY41

Did you have message or image capture

Try test use without use GPU,  may be GPU problem

What is you GPU?

 
I have idea for solve puzzle #64

up to we want to solve puzzle by anyone or keep it for solve puzzle by yourself

if we want to keep it for solve puzzle by yourself don't do anything every one find way yourself (may be next year some one on GPU mining farm try to solve it)

or if want to solve by give away to anyone it easy to solve with in 1 month (30 days)  just want to challenge over old records or not

just idea please ignore
my old idea make bitcrack version #64 build-in public pool (use free firebase database) and promote every one to know will be solve by millions people do it
make like 64bit pool do it but pool split share to who help to highest scan have record keyspace that scan already on database
but this idea is have database record pool but give away to who lucky can scan found this project call bitcrack LOTTO
it like a LOTTO for bitcoin value $28,797.44 USD (price for now post) I believe people over million will try to scan it by bitcrack special version puzzle #64
million people will scan free and one person lucky will win this lotto and get the key . user easy just run bitcrack (lotto) and waiting for lucky. it is FREE lotto no cost to pay just run computer
don't worry still have 78 address leave to solve  (next puzzle 66 may be can use this idea but for next #67 may be too high to do it)
may be add script to make transaction when found key and split for developer 30% (0.2) and who lucky win 70% (0.44)

for me I think leave that idea for every one find the way to crack it themselves  

puzzle is activity for fun
if people think waste of time , no body to crack it, it is boring (no excite)

may be puzzle #64 will last one to can crack because it is to high bits

Can someone please help me. The cubitcrack hangs after loading address and does nothing.
Thank you.

Because most see it as a waste of time/next to impossible, especially since everyone is trying to attack #64 solo.  #64 requires many resources to crack, and the next one, #66 will be 4  times as hard, and will get harder and harder as you go up.

EDIT: also, because there are other programs that can run multi gpu's on one instance, bitcrack is sort of archaic.  Maybe make it multi gpu and you can gain some new interest.