Large Bitcoin Collider (Collision Finders Pool)

[digaran]

Have you ever find any address with even a few satoshi in it? but lets assume that someday some one lucky enough finds one collision then what? next thing is going to take a few months or even years? this is like gathering thousands of keys and going on the street and test them one by one on every car and house to see if one of them could open a door to steal the car or steal from the house.

[1713576113]

1713576113

[1713576113]

1713576113

[1713576113]

1713576113

[1713576113]

1713576113

[1713576113]

1713576113

[becoin]

Soon the pool will have generated 500 tn keys and thus searched 1 quadrillion addresses. A complete and exhaustive search. I find that amusing too because I believe it hasn't been done before.

So, what? You haven't found a single collision! Oh, wait... you don't do it for the sake of finding collisions, right?

[Jude Austin]

Still no AMD generator?

[rico666]

Still no AMD generator?

I believe what we have now will also run on AMD, but it's untested. Why don't you send me your diagnostics-OpenCL.txt file?


I found another prformance optimization I'm working on right now. The performance problems experienced on AWS GPU instances may be related to that.
It has to do with startup times and was hidden from me, because of my 2.5 GB/s SSD...



Rico

[rico666]

Have you ever find any address with even a few satoshi in it? but lets assume that someday some one lucky enough finds one collision then what? next thing is going to take a few months or even years? this is like gathering thousands of keys and going on the street and test them one by one on every car and house to see if one of them could open a door to steal the car or steal from the house.

Actually we found two of them. One had 789900 and one 10000 satoshi "in it". https://lbc.cryptoguru.org/trophies
Maybe you would like to cover the basics by reading the thread before re-chewing what has been discussed before.


Rico

[Jude Austin]

Still no AMD generator?

I believe what we have now will also run on AMD, but it's untested. Why don't you send me your diagnostics-OpenCL.txt file?


I found another prformance optimization I'm working on right now. The performance problems experienced on AWS GPU instances may be related to that.
It has to do with startup times and was hidden from me, because of my 2.5 GB/s SSD...



Rico

Struggling to install AMD OpenCL on Ubuntu 14.04.4

Typically I just install FGLRX and it works fine.

Duh, helps if I actually installed Ubuntu instead of running it from the USB stick.

[becoin]

Actually we found two of them.

Actually you didn't.
I'm afraid you have to re-chew what's the difference between finding a private key and finding a collision?

Finding a private key doesn't prove anything as it can be result of:
- using a crappy RNG
- using a brain wallet with password like "Joe123"
- presenting your own private key as the one you've found

Lets assume you've found a private key. How will you prove it isn't one of your own private keys?

[rico666]

Actually we found two of them.

Actually you didn't.
I'm afraid you have to re-chew what's the difference between finding a private key and finding a collision?

Actually we did. The original question I answered was:

Have you ever find any address with even a few satoshi in it?

So yes, we found two of these. I'm afraid you have to re-chew basic education, including text understanding.
Thanks for keeping the thread alive, though.

And:

Lets assume you've found a private key. How will you prove it isn't one of your own private keys?

Seriously? For the 3rd time? Are you a troll, one account to many people, suffering from amnesia?
Really becoin, please seek professional help.


Rico

[becoin]

Seriously? For the 3rd time?

And for the 3rd time what I get is...

basic education
text understanding
keeping the thread alive
you a troll
one account to many people
suffering from amnesia
seek professional help

... all the usual stuff that people spit when they lack arguments!

[rico666]

... all the usual stuff that people spit when they lack arguments!

Yes, I admit: I'm lacking the arguments here.
Maybe someone else would like to try, because I have some coding and testing and releases to do this weekend.
So you managed again to suck 10 minutes of that precious time, but no more.


Rico

[cz3kit]

... all the usual stuff that people spit when they lack arguments!

Yes, I admit: I'm lacking the arguments here.
Maybe someone else would like to try, because I have some coding and testing and releases to do this weekend.
So you managed again to suck 10 minutes of that precious time, but no more.


Rico

Don't worry about him and ignore him. I guess that's the best you can do. You can't argue with people who do not have your mindset. You are interested to see if there is a possibility to create a collision - for the fun and interest. I would call it engineering work

Some people are not that way and he is one of them. He beliefs that this won't work and if you find something, well, then he considers you as a thief (my assumption).

[rico666]

Ok guys - this is it:

Everyone who has a working OpenCL install, is authorized for GPU, has an AVX2 capable CPU (I will provide sse42 GPU clients later) and feels bold enough, please download the beta from.

https://lbc.cryptoguru.org/downloads/beta/LBC

Will the GPU client work in a VM? No. So if any of this is not met, please do not bother right now, you'd only find pain and sorrow.

Attention! This LBC client is too advanced for your CPU generator (except you have a Skylake CPU and its newest generator)
so keep your older LBC somewhere or be prepared to delete this new one if you plan to use a CPU client again.

Maybe you would like to have your GPU experiment in another folder. Let's say you have a folder "collider", so copy it

Code:

$ cp -a collider gcollider

In the folder gcollider remove bench.pst, LBC and your gen-hrdcore-avx2-linux64 and eventually a present diagnostics-OpenCL.txt
and put the beta LBC in that folder

Code:

$ cd gcollider
$ rm -f bench.pst LBC gen-hrdcore-avx2-linux64 diagnostics-OpenCL.txt 
$ wget https://lbc.cryptoguru.org/downloads/beta/LBC
$ chmod a+x LBC

Ok, hour of truth:

upon 1st invocation you should see:

Code:

$ ./LBC --gpu -t 1 -l 0
OpenCL diagnostics written.
GPU authorized: yes
OpenCL program written.

LBC will end here and you'll see two new files in your directory (diagnostics-OpenCL.txt and hash160_deparsed.cl) you keep both of them where they are.

upon the 2nd invocation (and all subsequent), you should fly:

Code:

$ ./LBC --gpu -t 1 -l 0
GPU authorized: yes
Best generator chosen: gen-hrdcore-gpu-linux64
New generator found. (DL-size: 0.72MB)
Benchmark info not found - benchmarking... done.
Your maximum speed is 2098648 keys/s per CPU core.
Ask for work... got blocks [416958585-416959096] (536 Mkeys)
oooooooooooooooooooooooooooooooo (8.06 Mkeys/s)

In theory.   If you remove the -l 0, it will not end after one iteration and you may want to ramp up the time, so

Code:

$ ./LBC --gpu -t 5
GPU authorized: yes
Best generator chosen: gen-hrdcore-gpu-linux64
Ask for work... got blocks [419776761-419779192] (2550 Mkeys)

Nice numbers - eh?


Also in the news, user manual under https://lbc.cryptoguru.org/man/user extended with topics about Id, security, setting BTC address, GPU and defaults-file.

check back often, I will update this as we go.


Rico

[SlarkBoy]

after following your instruction


1st run

Code:

root@RetardPunk:~/LBCGPU# ./LBC --id JXXXXX --secret XXXXXXXXXX --gpu -t 5 -l 1
GPU authorized: yes
Will use 4 CPUs.
Best generator chosen: gen-hrdcore-gpu-linux64
New generator found. (DL-size: 0.72MB)
Benchmark info not found - benchmarking... ./gen-hrdcore-gpu-linux64: /usr/local/cuda-8.0/lib64/libOpenCL.so.1: no version information available (required by ./gen-hrdcore-gpu-linux64)
Illegal instruction (core dumped)
done.
Your maximum speed is 22471793 keys/s per CPU core.
Ask for work... Server doesn't like us. Answer: toofast.

2nd run

Code:

root@RetardPunk:~/LBCGPU# ./LBC --id JXXXXX --secret XXXXXXXXXX --cpus 8 --gpu -t 5 -l 1
GPU authorized: yes
Best generator chosen: gen-hrdcore-gpu-linux64
Ask for work... Server doesn't like us. Answer: toofast.

run with -x

Code:

root@RetardPunk:~/LBCGPU# ./LBC --id JXXXXX --secret XXXXXXXXXX --cpus 8 --gpu -x
GPU authorized: yes
Best generator chosen: gen-hrdcore-gpu-linux64
Testing mode. Using page 0, turning off looping.
Benchmark info not found - benchmarking... ./gen-hrdcore-gpu-linux64: /usr/local/cuda-8.0/lib64/libOpenCL.so.1: no version information available (required by ./gen-hrdcore-gpu-linux64)
Illegal instruction (core dumped)
done.
Your maximum speed is 64860036 keys/s per CPU core.
./gen-hrdcore-gpu-linux64: /usr/local/cuda-8.0/lib64/libOpenCL.so.1: no version information available (required by ./gen-hrdcore-gpu-linux64)

[TooDumbForBitcoin]

When the International Quantum Computing Society inaugurates it's Cryptocurrency Hacker of the Year award in 2025, will it be called the Rico in honor of the pioneer in collision computing?

[rico666]

after following your instruction


1st run

Code:

Illegal instruction (core dumped)

And we are sure the CPU is AVX2 capable?


Rico

[rico666]

When the International Quantum Computing Society inaugurates it's Cryptocurrency Hacker of the Year award in 2025, will it be called the Rico in honor of the pioneer in collision computing?

I don't know. Quantum computing is only 4th on my TODO list.

1. Full-GPU client
2. FPGA implementation
3. ASIC
4. Quantum Computing


Rico

[shifty252]

So, the beta is available to top30 and paying users? or for everyone with OpenCL installed?

[rico666]

So, the beta is available to top30 and paying users? or for everyone with OpenCL installed?

OpenCL is necessary but not sufficient.
The beta (and then the production client) is available to everyone, but it wikl work only for users who have a gpuauth: 1 attached to their id.

How does that gpuauth=1 happen?

• You are a top30 Gkeys contributor
• You fork out 0.1 BTC - which are distributed among all pool members (except me) who have a BTC address set
• You get a gpuauth set to 1 by decree (i.e. for special services)

Rico

[shifty252]

So, the beta is available to top30 and paying users? or for everyone with OpenCL installed?

OpenCL is necessary but not sufficient.
The beta (and then the production client) is available to everyone, but it wikl work only for users who have a gpuauth: 1 attached to their id.

How does that gpuauth=1 happen?

• You are a top30 Gkeys contributor
• You fork out 0.1 BTC - which are distributed among all pool members (except me) who have a BTC address set
• You get a gpuauth set to 1 by decree (i.e. for special services)

Rico

Thanks for clarifying. Top30 here I come, unless I stumble upon 0.1 BTC....

[shifty252]

aaand I got blacklisted while using the beta build only on CPU.

anyway to lift the ban so i can get back to regular LBC ?