Large Bitcoin Collider Thread 2.0

[Hereticalsauce]

Please update the LBC performance spreadsheet with your keys/sec instead of this thread:
https://docs.google.com/spreadsheets/d/1n6rh-0fMVYPEd69cD-3YPcFlgJaxBNh_bZr94kQSvIs/edit?usp=sharing

[1714052628]

1714052628

[adaseb]

Ok, I finally sold those Bitcore AND Bitcoin Gold coins from the 54th key.

I think I sold the BTG for around ~0.038 and the Bitcore was like $8 or so. So I will send over half as promised.

However little did I know that Hitbtc sends BTC from Segwit addresses and my Electrum isn't upgraded yet. So I need to upgrade first before I can send over the BTC to you. So give me a few days to do so.

[adaseb]

Ok done. Here is the transaction ID

https://blockchain.info/tx/1ec00c68cf375cec164271086cae7e8897302da299c7a6de5038f586c4b95823


Lets find the 55th number now!

[itod]

Excuse me for not reading everything about LBC, but who found the keys corresponding to 0.161 - 0.256 BTC outputs range in the Puzzle transaction? There are somehow derived from 0.001 to 0.054 private keys?

[BurtW]

Excuse me for not reading everything about LBC, but who found the keys corresponding to 0.161 - 0.256 BTC outputs range in the Puzzle transaction? There are somehow derived from 0.001 to 0.054 private keys?

Those were moved by the creator (of the puzzle) because they were redundant.

[arulbero]

Excuse me for not reading everything about LBC, but who found the keys corresponding to 0.161 - 0.256 BTC outputs range in the Puzzle transaction? There are somehow derived from 0.001 to 0.054 private keys?

https://bitcointalk.org/index.php?topic=1306983.msg18765941#msg18765941

[itod]

Ah, thanks for clarification. It's obvious when you see the answer

[holy_ship]

I do understand that the author needed to implement a way to make sure the data sent to the server is always valid. If not, the whole project is failing...

Hello, guys! Just joined my modest 10Mkps to your great project. But I still have a couple of questions...

1. What perl program does when the collision is found? Only writes to file FOUND.txt? If I someone overlook/delete this file, nobody will even know that we found the key and the worst thing - this range should be marked as empty? 

2. Is there any validation of results? Or just checking checksums of the program itself? Why not to compute each block's checksum (even simplest in couple of bytes) and send to server?
Then make 10% of work-units overlap. If checksums do not match, sending to third client and marking ALL work from client with wrong checksum as undone and ban him if there were several badly processed blocks.

For now it seems like we are sifting with too large sieve? 

[rico666]

1. What perl program does when the collision is found? Only writes to file FOUND.txt? If I someone overlook/delete this file, nobody will even know that we found the key and the worst thing - this range should be marked as empty? 

See https://lbc.cryptoguru.org/man/user#hooks

2. Is there any validation of results? Or just checking checksums of the program itself? Why not to compute each block's checksum (even simplest in couple of bytes) and send to server?
Then make 10% of work-units overlap. If checksums do not match, sending to third client and marking ALL work from client with wrong checksum as undone and ban him if there were several badly processed blocks.

For now it seems like we are sifting with too large sieve? 

Some overlap happens already, although it's less than 2%
Work is also being re-issued.

Not sure what you mean by the "too large sieve". Rest assured, validation is working fine.

[timisis]

Joining up!

[arulbero]

A number of people have raised this matter: how to read properly the FOUND.txt file?

I decided to share a little python script.


First we generate a FOUND.txt file to make a test:

Code:

$ ./LBC -x
Testing mode. Using page 0, turning off looping.
Benchmark info not found - benchmarking... done.
Your speed is roughly ............ keys/s per CPU core.
o
Test ok. Your test results were stored in FOUND.txt.
Have a look and then you may want to remove the file.

$ ls
FOUND.txt
....

$ more FOUND.txt 
2d17543d32448acc7a1c43c5f72cd5be459ab302:u:priv:0000000000000000000000000000000000000000000000000000000000000001+0x5e
02e62151191a931d51cdc513a86d4bf5694f4e51:c:priv:0000000000000000000000000000000000000000000000000000000000000001+0x65
9d74ffdb31068ca2a1feb8e34830635c0647d714:u:priv:00000000000000000000000000000000000000000000000000000000000f9001+0xf8c
3d6871076780446bd46fc564b0c443e1fd415beb:c:priv:00000000000000000000000000000000000000000000000000000000000f9001+0xf8c

Then you can use this python script, called "lbc_output.py": https://www.dropbox.com/s/q1sgc4gbb26vc99/lbc_output.py?dl=0

Copy the line of FOUND.TXT you are interested of and you get the result:

Code:

$ ./lbc_output.py 2d17543d32448acc7a1c43c5f72cd5be459ab302:u:priv:0000000000000000000000000000000000000000000000000000000000000001+0x5e

Private key : 000000000000000000000000000000000000000000000000000000000000005f
Public  key :
          x : 15d9441254945064cf1a1c33bbd3b49f8966c5092171e699ef258dfab81c045c
          y : d56eb30b69463e7234f5137b73b84177434800bacebfc685fc37bbe9efe4070d
 
PrKey WIF u.: 5HpHagT65TZzG1PH3CSu63k8DbpvD8s5ip4nEB3kEsreMQiR4w7
Address u.  : 2d17543d32448acc7a1c43c5f72cd5be459ab302
Address u.  : 157RMZhbBLC1wucv3jxQqqHjbKezL1yy7g

What does the script?

First it reads and parses the line.
Then it computes the private key (it does the addition, in our example: 0000000000000000000000000000000000000000000000000000000000000001 + 0x5e) and using the ecc arithmetic it generates the public key. Then it generates the address (compressed or uncompressed) and checks if it matches with the address in FOUND.txt (in this case 2d17543d32448acc7a1c43c5f72cd5be459ab302).

Finally it provides the private key in WIF format and the address b58 encoded.

[holy_ship]

By the way, don't you think that starting from very beginning is a bad idea?

I guess most of us are hoping to discover abandoned wallets from early 2010-s.
At that times harsh cryptofreaks didn't have fancy bitcoin-core software and generated private keys themselves.
So, they at least have seen the result of RNG.
And the secret key with 40 leading zeros looks very stupid.
Yeah, I understand that any value from RNG has the same probability to occur, but still...

Why not to move to middle of 256-bit range (or 2^160 - I'm not sure what exactly we are bruteforcing)?

[holy_ship]

See https://lbc.cryptoguru.org/man/user#hooks
Not sure what you mean by the "too large sieve".

Thanks for the answer, rico666! English is not my native language.
It seems to me that LBC now has rather weak control of clients (correctness of their checks and return of results).
So the effect is: in list of Trophies we see far less records than it should be.

Also the main sense and benefit of pool is sharing award between all members (proportionally to contribution).
But now the single member takes everything (and can even not report the find!).

Actually, I do not see any advantages of being member of LBC versus solo mode.
Yeah, I've read the FAQ, but I totally disagree with the answer.
The range of LBC is known (and I think it is not best).
Solo "miner" can easily use another range without being a member of LBC.

The point of this message is not "hey, everyone, run from LBC", but to make LBC better and possibly more popular.

[rico666]

See https://lbc.cryptoguru.org/man/user#hooks
Not sure what you mean by the "too large sieve".

Thanks for the answer, rico666! English is not my native language.
It seems to me that LBC now has rather weak control of clients (correctness of their checks and return of results).
So the effect is: in list of Trophies we see far less records than it should be.

Also the main sense and benefit of pool is sharing award between all members (proportionally to contribution).
But now the single member takes everything (and can even not report the find!).

Actually, I do not see any advantages of being member of LBC versus solo mode.
Yeah, I've read the FAQ, but I totally disagree with the answer.
The range of LBC is known (and I think it is not best).
Solo "miner" can easily use another range without being a member of LBC.

The point of this message is not "hey, everyone, run from LBC", but to make LBC better and possibly more popular.

No hard feelings, but I think you have still a lot to read and learn about the project and the concept and tech behind it.

Most of your "suggestions" are result of being badly informed - really.

[holy_ship]

rico666 thanks for not sending me to GTFO  Tried to read "What We Do" more attentive.

One question still remains: did you estimate required funds to make FPGA or ASIC clients for LBC? I guess ordinary bitcoin mining ASICs wouldn't be effective (or suitable at all)?

[rico666]

rico666 thanks for not sending me to GTFO  Tried to read "What We Do" more attentive.

One question still remains: did you estimate required funds to make FPGA or ASIC clients for LBC? I guess ordinary bitcoin mining ASICs wouldn't be effective (or suitable at all)?

ASICs - as the "Application-Specific" part suggests - are of use only for a very narrow usage case. Bitcoin miners have scrapyard value beyond BTC* mining.

I asked http://www.orsoc.se/ what the cost of ASIC development might be about 2 months after launching the pool.
ORSoC is/was the technology development company behind KnC and I thought it would be a good address to ask.

Unfortunately never got any answer. I might try again.

Bitcoin miners do a SHA256d essentially.
The LBC does  ECC, followed by hash160, followed by bloom filter lookup

Now the hash160 is actually less demanding than SHA256d and if it was only for that, I'm pretty sure that hash160 ASICs could deliver more performance than SHA256d ASICs.
The ECC however, requires 256bit multiplications and that is serious stuff taking up whole FPGA circuits. (at least until a few years ago)

At the moment we do these ECC things on CPU and hash160 on GPU

An ASIC doing hash160 and Bloom-filter lookup (512MB chips containing the BLF connected directly to the ASIC) I can imagine, but I can't imagine how to feed it with ECC data.
ECC on ASIC I can't imagine, but then again, I am a VHDL, Verilog, FPGA, ASIC noob.

It would certainly be good if we managed to establish at least some informal smalltalk with ORSoC (or similar) engineers about this.

[rico666]

We have now our own directory.io, clone as the original site seems to have gone.

Clicking on the directory.io in https://lbc.cryptoguru.org/stats leads to https://lbc.cryptoguru.org/dio/101091723026432 (or whatever the page at the time of the click is)

I've enabled a rate limiter for all my beloved leechers out there, who still haven't understood how the page works.

Then you can use this python script, called "lbc_output.py": https://www.dropbox.com/s/q1sgc4gbb26vc99/lbc_output.py?dl=0

Copy the line of FOUND.TXT you are interested of and you get the result:

Code:

$ ./lbc_output.py 2d17543d32448acc7a1c43c5f72cd5be459ab302:u:priv:0000000000000000000000000000000000000000000000000000000000000001+0x5e

Private key : 000000000000000000000000000000000000000000000000000000000000005f
Public  key :
          x : 15d9441254945064cf1a1c33bbd3b49f8966c5092171e699ef258dfab81c045c
          y : d56eb30b69463e7234f5137b73b84177434800bacebfc685fc37bbe9efe4070d
 
PrKey WIF u.: 5HpHagT65TZzG1PH3CSu63k8DbpvD8s5ip4nEB3kEsreMQiR4w7
Address u.  : 2d17543d32448acc7a1c43c5f72cd5be459ab302
Address u.  : 157RMZhbBLC1wucv3jxQqqHjbKezL1yy7g

With the above, the reverse way is possible too:

https://lbc.cryptoguru.org/dio/priv/5HpHagT65TZzG1PH3CSu63k8DbpvD8s5ip4nEB3kEsreMQiR4w7

will take you to the page containing that privkey and the address.

[andreyyv]

Hi! what is the -c in process? why they are diferent? 
http://c2n.me/3Qpu2Gr

[rico666]

Hi! what is the -c in process? why they are diferent? 

Client -> Generator challenge. To make sure the client is talking to a legit generator.

[smmintl]

Total newb here when it comes to linux but please take a look at the attached image and let me know if I actually started the collider

https://i.imgur.com/BLSDMjc.png

If it's started how can I see the FOUND.txt file in case something is found? Any suggestion is much appreciated.

edit: also, does 1,8 Mkeys/s mean that I'm searching through 1,800,000 keys per second?