Bitcoin puzzle transaction ~32 BTC prize to who solves it

[Feron]

ok, but it's still awesome, only if the kangaroo worked according to my ideas, it would be a different coffee

[1714611391]

1714611391

[1714611391]

1714611391

[1714611391]

1714611391

[Tepan]

are there any comparisons available in relation to speed and success between BSGS-CUDA vs. KANGAROO-CUDA ?

for me, it's faster BSGS CUDA

[citb0in]

for me, it's faster BSGS CUDA

is there any existing BSGS-CUDA linux version available which I could install and try ? I cannot find anything on github. I found a BSGS-CUDA which is Windows only and even then you would need a licensed PureBasic installation for running it. Any linux build available ?

[Tepan]

for me, it's faster BSGS CUDA

is there any existing BSGS-CUDA linux version available which I could install and try ? I cannot find anything on github. I found a BSGS-CUDA which is Windows only and even then you would need a licensed PureBasic installation for running it. Any linux build available ?

I never found the BSGS-CUDA running on linux/debian, still use the Windows version, There's the github you can download the BSGS-CUDA windows and already compiled.
you just need extract the zip, and put the exe on the same file.

test your luck here, it's safe no random malware.

https://github.com/Etayson/BSGS-cuda/releases/tag/v1.7.3

[EDIT]


I try to scan the #65 Puzzle with BSGS-CUDA, using GTX1650Ti [Testing the 30 # puzzle]

results,

START RANGE= 0000000000000000000000000000000000000000000000000000000037d72a06
  END RANGE= 000000000000000000000000000000000000000000000000000000003fffffff
WIDTH RANGE= 000000000000000000000000000000000000000000000000000000000828d5f9
SUBpoint= (2bdeb0a84d7f1cf940ab6a47875939fcfb52e07e8598de29978f67b3979dc202, a0c47daf4ac0e02a1f3754e25bcd471d9a599db7432d80794df604814fcf75fd)
Save work every 180 seconds

FINDpubkey: 030d282cf2ff536d2c42f105d0b8588821a915dc3f9a05bd98bb23af67a2e92a5b
Cnt:1 [1][ 0 ] = 0 MKeys/s x2^27=2^27.00 t:00:00:00
KEY[1]: 0x000000000000000000000000000000000000000000000000000000003d94cd64
   Pub: 030d282cf2ff536d2c42f105d0b8588821a915dc3f9a05bd98bb23af67a2e92a5b
Working time 00:00:00s
Total time 00:00:05s

GPU#0 job finished
GPU#0 thread finished
cuda finished ok

only 5 seconds.

IMO imagine the puzzle maker show pubkey the #66 Puzzle, maybe it's take the same time like 10-20 seconds haha.

[WanderingPhilospher]

for me, it's faster BSGS CUDA

is there any existing BSGS-CUDA linux version available which I could install and try ? I cannot find anything on github. I found a BSGS-CUDA which is Windows only and even then you would need a licensed PureBasic installation for running it. Any linux build available ?

I never found the BSGS-CUDA running on linux/debian, still use the Windows version, There's the github you can download the BSGS-CUDA windows and already compiled.
you just need extract the zip, and put the exe on the same file.

test your luck here, it's safe no random malware.

https://github.com/Etayson/BSGS-cuda/releases/tag/v1.7.3


I try to scan the #65 Puzzle with BSGS-CUDA, using GTX1650Ti

results,

START RANGE= 0000000000000000000000000000000000000000000000000000000037d72a06
  END RANGE= 000000000000000000000000000000000000000000000000000000003fffffff
WIDTH RANGE= 000000000000000000000000000000000000000000000000000000000828d5f9
SUBpoint= (2bdeb0a84d7f1cf940ab6a47875939fcfb52e07e8598de29978f67b3979dc202, a0c47daf4ac0e02a1f3754e25bcd471d9a599db7432d80794df604814fcf75fd)
Save work every 180 seconds

FINDpubkey: 030d282cf2ff536d2c42f105d0b8588821a915dc3f9a05bd98bb23af67a2e92a5b
Cnt:1 [1][ 0 ] = 0 MKeys/s x2^27=2^27.00 t:00:00:00
KEY[1]: 0x000000000000000000000000000000000000000000000000000000003d94cd64
   Pub: 030d282cf2ff536d2c42f105d0b8588821a915dc3f9a05bd98bb23af67a2e92a5b
Working time 00:00:00s
Total time 00:00:05s

GPU#0 job finished
GPU#0 thread finished
cuda finished ok

only 5 seconds.

IMO imagine the puzzle maker show pubkey the #66 Puzzle, maybe it's take the same time like 10-20 seconds haha.

Lol…what is this?! You ran a 30 bit range in 5 seconds?! Kangaroo can do that in 1 second with CPU only.
Where did you run 65 bit? And those results?
Y’all need to stop unless I’m reading everything wrong lol.

[Evillo]

for me, it's faster BSGS CUDA

is there any existing BSGS-CUDA linux version available which I could install and try ? I cannot find anything on github. I found a BSGS-CUDA which is Windows only and even then you would need a licensed PureBasic installation for running it. Any linux build available ?

I never found the BSGS-CUDA running on linux/debian, still use the Windows version, There's the github you can download the BSGS-CUDA windows and already compiled.
you just need extract the zip, and put the exe on the same file.

test your luck here, it's safe no random malware.

https://github.com/Etayson/BSGS-cuda/releases/tag/v1.7.3


I try to scan the #65 Puzzle with BSGS-CUDA, using GTX1650Ti

results,

START RANGE= 0000000000000000000000000000000000000000000000000000000037d72a06
  END RANGE= 000000000000000000000000000000000000000000000000000000003fffffff
WIDTH RANGE= 000000000000000000000000000000000000000000000000000000000828d5f9
SUBpoint= (2bdeb0a84d7f1cf940ab6a47875939fcfb52e07e8598de29978f67b3979dc202, a0c47daf4ac0e02a1f3754e25bcd471d9a599db7432d80794df604814fcf75fd)
Save work every 180 seconds

FINDpubkey: 030d282cf2ff536d2c42f105d0b8588821a915dc3f9a05bd98bb23af67a2e92a5b
Cnt:1 [1][ 0 ] = 0 MKeys/s x2^27=2^27.00 t:00:00:00
KEY[1]: 0x000000000000000000000000000000000000000000000000000000003d94cd64
   Pub: 030d282cf2ff536d2c42f105d0b8588821a915dc3f9a05bd98bb23af67a2e92a5b
Working time 00:00:00s
Total time 00:00:05s

GPU#0 job finished
GPU#0 thread finished
cuda finished ok

only 5 seconds.

IMO imagine the puzzle maker show pubkey the #66 Puzzle, maybe it's take the same time like 10-20 seconds haha.

You are using a wrong range. Puzz 65 is 17 characters not 8

Puzz 66 range in BSGS would take an hour  a few minutes at most.

----

Btw, is anyone trying their luck with puzz 125? The only discussion i see going here is about 66 lol. Looks like 125 bits is too big to give any kind of hope.

[teslamegapacks1919]

When playing with some iceland tools which is https://github.com/iceland2k14/kangaroo i got this speed on CPU
 

• [24894.64 TeraKeys/s][Kang 28672][Count 2^29.29/2^29.10][Elapsed 06s][Dead 1][RAM 53.4MB/46.0MB]

Is it good bad fake idk but like i know iceland tools are good

[Tepan]

for me, it's faster BSGS CUDA

is there any existing BSGS-CUDA linux version available which I could install and try ? I cannot find anything on github. I found a BSGS-CUDA which is Windows only and even then you would need a licensed PureBasic installation for running it. Any linux build available ?

I never found the BSGS-CUDA running on linux/debian, still use the Windows version, There's the github you can download the BSGS-CUDA windows and already compiled.
you just need extract the zip, and put the exe on the same file.

test your luck here, it's safe no random malware.

https://github.com/Etayson/BSGS-cuda/releases/tag/v1.7.3


I try to scan the #65 Puzzle with BSGS-CUDA, using GTX1650Ti

results,

START RANGE= 0000000000000000000000000000000000000000000000000000000037d72a06
  END RANGE= 000000000000000000000000000000000000000000000000000000003fffffff
WIDTH RANGE= 000000000000000000000000000000000000000000000000000000000828d5f9
SUBpoint= (2bdeb0a84d7f1cf940ab6a47875939fcfb52e07e8598de29978f67b3979dc202, a0c47daf4ac0e02a1f3754e25bcd471d9a599db7432d80794df604814fcf75fd)
Save work every 180 seconds

FINDpubkey: 030d282cf2ff536d2c42f105d0b8588821a915dc3f9a05bd98bb23af67a2e92a5b
Cnt:1 [1][ 0 ] = 0 MKeys/s x2^27=2^27.00 t:00:00:00
KEY[1]: 0x000000000000000000000000000000000000000000000000000000003d94cd64
   Pub: 030d282cf2ff536d2c42f105d0b8588821a915dc3f9a05bd98bb23af67a2e92a5b
Working time 00:00:00s
Total time 00:00:05s

GPU#0 job finished
GPU#0 thread finished
cuda finished ok

only 5 seconds.

IMO imagine the puzzle maker show pubkey the #66 Puzzle, maybe it's take the same time like 10-20 seconds haha.

Lol…what is this?! You ran a 30 bit range in 5 seconds?! Kangaroo can do that in 1 second with CPU only.
Where did you run 65 bit? And those results?
Y’all need to stop unless I’m reading everything wrong lol.

you too busy to judge my opinion bro, look at the WORKING TIME it's say 00:00:00s, i run this command took 5 second to DONE. you know i mean right ?
so why you just think it's so instant ? if this so fast? why you need to run the command manualy, so ?  
Kangaroo 1 second ?


Here's the right RANGE if you mention me about the wrong range.
i have more range nonce about the pubkey 65.

Results,

START RANGE  = 00000000000000000000000000000000000000000000000186053f25c09123e2
END RANGE     = 000000000000000000000000000000000000000000000001ffffffffffffffff
WIDTH RANGE= 00000000000000000000000000000000000000000000000068fae0d84f6eed2b
SUBpoint= (028e1a50d76d381a1354bee348ef60dc7b4ad43cae7eb446673f2771f9033cda, 99cd9a310590c57bfbfac6288ba9e9b340780a1ad2fdd0c1c7dc9498033e7074)
Save work every 180 seconds

FINDpubkey: 0230210c23b1a047bc9bdbb13448e67deddc108946de6de639bcc75d47c0216b1b
Cnt:f35000000000001 [1][ 375 ] = 375 MKeys/s x2^27=2^55.55 t:00:00:00
KEY[1]: 0x000000000000000000000000000000000000000000000001a838b13505b26867
   Pub: 0230210c23b1a047bc9bdbb13448e67deddc108946de6de639bcc75d47c0216b1b
Working time 00:00:00s
Total time 00:00:08s
GPU#0 job finished
GPU#0 thread finished
cuda finished ok

[yoshimitsu777]

When playing with some iceland tools which is https://github.com/iceland2k14/kangaroo i got this speed on CPU
 

• [24894.64 TeraKeys/s][Kang 28672][Count 2^29.29/2^29.10][Elapsed 06s][Dead 1][RAM 53.4MB/46.0MB]

Is it good bad fake idk but like i know iceland tools are good

iceland2k14 is proprietary and all his tools contain backdoor
do not use iceland2k14 libraries unless you want to loose your keys or findings
good luck

[AlanJohnson]

----

Btw, is anyone trying their luck with puzz 125? The only discussion i see going here is about 66 lol. Looks like 125 bits is too big to give any kind of hope.

I must admit i gave up seeking solutions for those puzzles. It consumes to much time and gives nothing.

I've tried 125 with BSGS and Kangaroo  ... i've tried 66 in random mode ... and my conclusions is: These numbers are just too big for seeking with home computer.

Even the topic name is misleading here - These are not PUZZLES ... there is no puzzle , no pattern , nothing. Just massive random numbers. With current hardware possibilities it could be solved but we need massive amonut of hardware to do it in a reasonable time.

It should be called something like "Computational Power Competition" or "Bitcoin Benchmark" not a "puzzle".

The only thing that can be solved by an average person are puzzles about maybe 90 - 100 bits but only in case the creator will reveal the public keys. But then solving puzzle 66 is a matter of few minutes and only competition is who will notice first that the public key is available ( and I doubt that was the goal of creator of all this thing).

When you realize with how big numbers we deal here and what is the speed of our machines the only solution is to give up.

Of course one would treat this as a lottery but considering the power consumption and ultra low chances to win  a (let's be honest) not that big reward (its significant amount but not a life changer) i doubt it's worth it...

Anyway it was useful experience from educational point of view.

[Lolo54]

The only thing that can be solved by an average person are puzzles about maybe 90 - 100 bits but only in case the creator will reveal the public keys. But then solving puzzle 66 is a matter of few minutes and only competition is who will notice first that the public key is available ( and I doubt that was the goal of creator of all this thing).

OP would have no interest in revealing the PubKey from #66 to #75 today it would be almost automatically discovered. The goal of OP was to demonstrate the solidity / resistance of BTC in the face of the ingenuity and the very power of a community!

for puzzle #66 if the pubkey was revealed it would take about 2s with an average
#67 4s / # 68 8s / # 69 16s / #71 1m etc of course subject to identical key location  and using average material

for #80 with average hardware kangaroo it takes 1 hour to find the privateKey .

[hugoranking]

Nice

[Evillo]

----

Btw, is anyone trying their luck with puzz 125? The only discussion i see going here is about 66 lol. Looks like 125 bits is too big to give any kind of hope.

I must admit i gave up seeking solutions for those puzzles. It consumes to much time and gives nothing.

I've tried 125 with BSGS and Kangaroo  ... i've tried 66 in random mode ... and my conclusions is: These numbers are just too big for seeking with home computer.

Even the topic name is misleading here - These are not PUZZLES ... there is no puzzle , no pattern , nothing. Just massive random numbers. With current hardware possibilities it could be solved but we need massive amonut of hardware to do it in a reasonable time.

It should be called something like "Computational Power Competition" or "Bitcoin Benchmark" not a "puzzle".

The only thing that can be solved by an average person are puzzles about maybe 90 - 100 bits but only in case the creator will reveal the public keys. But then solving puzzle 66 is a matter of few minutes and only competition is who will notice first that the public key is available ( and I doubt that was the goal of creator of all this thing).

When you realize with how big numbers we deal here and what is the speed of our machines the only solution is to give up.

Of course one would treat this as a lottery but considering the power consumption and ultra low chances to win  a (let's be honest) not that big reward (its significant amount but not a life changer) i doubt it's worth it...

Anyway it was useful experience from educational point of view.

From an individual point of view, it's almost impossible and is not worth it due to the extreme difficulty of the puzzles. But from a community perspective, it's really doable, after all, that's how puzz 64 was solved. If hundreds of ppl are searching, one of them will eventually land on the right key. Not to mention that we now know how long it would take one person with average resources to solve 66 bits of the gigantic 160 bits of Bitcoin's difficulty. Like you said, it's a great educational experience. I personally would romanticize this puzzle as it taught me a lot about cryptography, programming, math, linux and much more. Whoever created this thing not only made us test the security of Bitcoin to the bone, but also opened many doors for us to explore the world of cryptography and big numbers during our chase for the prize. He deserves a trophy.

[lordfrs]

Hello, I've been following this thread for a year. I stocked all of them, including many archives that were deleted from the internet, along with their source code.

I just came up with a new idea and I need a little help with it.

Do you have fast point extraction algorithm for python ?

Thanks..

Not: sorry for my bad english, i don't know english i use google translate

[Feron]

you don't need any magic code guess the first 6-7 hex numbers and scan the last 10 in bitcrack and voila you have a puzzle win in 5 minutes ddd

[lordfrs]

you don't need any magic code guess the first 6-7 hex numbers and scan the last 10 in bitcrack and voila you have a puzzle win in 5 minutes ddd

I'm writing an algorithm for all addresses with pubkey exposed. This is impossible with your method.

[Feron]

You can try, but don't expect miracles, almost all public keys have 256 bit private keys, which is a different universe, simply put, a waste of time

[Kamoheapohea]

If you guys try to skip non-random keys (like those containing repeating patterns like 111) then you can also try to skip ranges when some prefixing characters of the address were found.
Example: You found a prefix 1XXXXX of target 1XXXXX... so you can skip ahead like nBits because it is unlikely 1XXXXX will appear in nBits again. The skipped offset nBits should be very conservative/pessimistic.

Or when doing it multithreaded you can mark nBits around the found prefix 1XXXXX as "unlikely" and continue the search only in the "possible" ranges. You could decrease the unlikely nBits as you progress to eventually realize this does not work too.

[Tepan]

I just do some math with #66 Puzzle.

You just need bigA$$ Mkeys. (e.g 5000 Mkeys)

time = (range size) / (speed * 10^6)

Where range size is the number of keys in the range (3ffffffffffffffff - 20000000000000000 + 1 = 3ffffffffffffffe1), speed is the speed of the search in millions of keys per second, and 10^6 is used to convert from millions to individual keys per second.

Time = (3ffffffffffffffe1) / (5000 * 10^6) = 7671.29 seconds

Therefore, it would take approximately 7671.29 seconds (or about 2.13 hours) to search the given range at a speed of 5000 Mkeys per second.

Device i use was 1650Ti and give 343 Mkeys speed with Keyhunt Albertobsd, 15X GPUs needed..

Time to win the lottery haha  

this method only Kangaroo and hash160.

sorry for not mention about the 66 bit range haha, i already double check on reply below

but i feelin weird about the keybit range about 66 itself. its maybe arround 20000000000000000:4ffffffffffffffff not 3.  

this just IMO , consider to bullyin my opinion you better skip to read this.
#CMIIW

[Tepan]

There's the math comes again,

66 Bit puzzle  


the total number of keys in this range can be calculated by taking 16 to the power of 17 (i.e., 16^17), which is equal to 295,147,905,179,352,825,856.

So, there are a total of 295,147,905,179,352,825,856 keys in the given range.

If you have a speed of 343 million keys per second (Mkeys/s), the result's,

Total number of keys = 295,147,905,179,352,825,856
Time = Total number of keys / Speed

Substituting the values, we get:

Time = 295,147,905,179,352,825,856 / 343,000,000
Time = 859,488,566,568 seconds

Therefore, it would take approximately 859,488,566,568 seconds (or about 27,247 years) to generate all the possible keys in the given range at a speed of 343 million keys per second. This is a very long time and may not be practically feasible.

what if i have 5000 mkeys to scan those number of key ?
it would take approximately 59,029,581,036 seconds (or about 1,872 years) to generate all the possible keys in the given range at a speed of 5000 million keys per second, still lame haha  


and there's other scenario to win the ticket haha

For a 66-bit puzzle, the key range would be from 0 to 2^66 - 1, which is:

0 to 7,922,816,251,426,433

The number of keys in this range is:

2^66 = 73,786,976,294,838,206,464

At a scanning speed of 343 Mkey/s, the time it would take to scan this entire range would be:

Time = Number of Keys / Keys per Second

Time = (2^66) / 343000000 = 215625224 seconds
= 3593754 minutes
= 59896 hours
= 2495 days
≈ 6.84 years

So, it would take approximately 6.84 years to scan the entire key range of a 66-bit puzzle at the given scanning speed.

If we were to randomly hit the private key in the middle of the range, then we would only need to scan half of the key range, which is:

2^65 = 36,893,488,147,419,103,232

Using the same formula as above, the time it would take to scan half of the key range would be:

Time = (2^65) / 343000000 = 107812612 seconds
= 1796877 minutes
= 29948 hours
= 1248 days
≈ 3.42 years

So, if we randomly hit the private key in the middle of the key range, it would take approximately 3.42 years to scan half of the key range at the given speed of 343 Mkey/s.


btw, my close friend is neat, he's building something on ASICs L3+ to give a proof to scan the puzzle for mean time haha.
let's see..