Bitcoin puzzle transaction ~32 BTC prize to who solves it

[giorgi1919]

Hello, Bitcointalk community,

Link To Source Code https://github.com/georgetoloraia/intelligent-puzzle-solver

I have been fascinated by the cryptographic puzzles within our space and have developed an open-source, educational C++ project aimed at the 71-bit puzzle range (2^70 to 2^71). This is not a brute-force tool but an adaptive intelligence system that learns from its search attempts to optimize the process.

This solver implements a multi-threaded, self-optimizing search strategy. Instead of blind sequential scanning, it analyzes partial hash160 collisions to intelligently guide its search.

Core Concept: When a generated key produces a hash160 that matches the first N characters of the target, the system identifies the key's region as "promising" and focuses more computational power around it.

Self-Evolving Intelligence: The program dynamically adjusts its search threshold and resource allocation in real-time based on what it learns.

If anyone is interested, please help me improve, I love all ideas, even the smallest ones.

Do you know that what you did is just a little basic pre-filtering before a search?  

"Virtuose" -> You're right... I'll try to modify it... I wanted filtering while working, thanks.

Code lovers. Take a look at the logic, give me some feedback..

I made a few corrections
I know it will be faster with AVX and AVX2. But I can't afford that much ))

[SRG02289]

477007882A8842614B   1PWo3JeB9jTUxadvYh1wXCYUpH2CsyD27E             
477009EA4543BFFAFB   1PWo3JeB9jtL4RK2wpYczGygdkn3v4ZjcR
47700B794FABB04B8A   1PWo3JeB9jK4jourU5YGZYEtnAKDsNwYzQ   
477011D6C809817D9F   1PWo3JeB9jwaCQkp9rvKDNd4QaXr5Jrh5S

[Virtuose]

477007882A8842614B   1PWo3JeB9jTUxadvYh1wXCYUpH2CsyD27E             
477009EA4543BFFAFB   1PWo3JeB9jtL4RK2wpYczGygdkn3v4ZjcR
47700B794FABB04B8A   1PWo3JeB9jK4jourU5YGZYEtnAKDsNwYzQ   
477011D6C809817D9F   1PWo3JeB9jwaCQkp9rvKDNd4QaXr5Jrh5S

Yeah, it's a lot of 1PWo3JeB9j all over the range 400000000000000000:7fffffffffffffffff 

[SRG02289]

477007882A8842614B   1PWo3JeB9jTUxadvYh1wXCYUpH2CsyD27E             
477009EA4543BFFAFB   1PWo3JeB9jtL4RK2wpYczGygdkn3v4ZjcR
47700B794FABB04B8A   1PWo3JeB9jK4jourU5YGZYEtnAKDsNwYzQ   
477011D6C809817D9F   1PWo3JeB9jwaCQkp9rvKDNd4QaXr5Jrh5S

Yeah, it's a lot of 1PWo3JeB9j all over the range 400000000000000000:7fffffffffffffffff 

Yes. We've already checked the entire range from 40x to 477x. And there are a lot of them... here's the latest:
4778AF41E8842B0652   1PWo3JeB9jbdqjJP6TYhNiweRjfeMoCfhv

[Frequence]

477007882A8842614B   1PWo3JeB9jTUxadvYh1wXCYUpH2CsyD27E             
477009EA4543BFFAFB   1PWo3JeB9jtL4RK2wpYczGygdkn3v4ZjcR
47700B794FABB04B8A   1PWo3JeB9jK4jourU5YGZYEtnAKDsNwYzQ   
477011D6C809817D9F   1PWo3JeB9jwaCQkp9rvKDNd4QaXr5Jrh5S

Yeah, it's a lot of 1PWo3JeB9j all over the range 400000000000000000:7fffffffffffffffff 

Yes. We've already checked the entire range from 40x to 477x. And there are a lot of them... here's the latest:
4778AF41E8842B0652   1PWo3JeB9jbdqjJP6TYhNiweRjfeMoCfhv

What’s your scanning speed?! G/s ?

[analyticnomad]

477007882A8842614B   1PWo3JeB9jTUxadvYh1wXCYUpH2CsyD27E             
477009EA4543BFFAFB   1PWo3JeB9jtL4RK2wpYczGygdkn3v4ZjcR
47700B794FABB04B8A   1PWo3JeB9jK4jourU5YGZYEtnAKDsNwYzQ   
477011D6C809817D9F   1PWo3JeB9jwaCQkp9rvKDNd4QaXr5Jrh5S

Yeah, it's a lot of 1PWo3JeB9j all over the range 400000000000000000:7fffffffffffffffff  

Yes. We've already checked the entire range from 40x to 477x. And there are a lot of them... here's the latest:
4778AF41E8842B0652   1PWo3JeB9jbdqjJP6TYhNiweRjfeMoCfhv

How many keys did you find? (1PWo3JeB9j) Probably around 50,000? What program do you use to search?

477007882A8842614B   1PWo3JeB9jTUxadvYh1wXCYUpH2CsyD27E             
477009EA4543BFFAFB   1PWo3JeB9jtL4RK2wpYczGygdkn3v4ZjcR
47700B794FABB04B8A   1PWo3JeB9jK4jourU5YGZYEtnAKDsNwYzQ   
477011D6C809817D9F   1PWo3JeB9jwaCQkp9rvKDNd4QaXr5Jrh5S

Yeah, it's a lot of 1PWo3JeB9j all over the range 400000000000000000:7fffffffffffffffff  

Yes. We've already checked the entire range from 40x to 477x. And there are a lot of them... here's the latest:
4778AF41E8842B0652   1PWo3JeB9jbdqjJP6TYhNiweRjfeMoCfhv

What’s your scanning speed?! G/s ?

It seems like it has a speed of 11,500,000,000,000 per second.

How is that speed even possible with deterministic scanning and w/out a pub key?

Splitting ranges w/multiple GPU's?

[SRG02289]

Splitting ranges w/multiple GPU's?

yes. time spent 5-6 months

[brainless]

Splitting ranges w/multiple GPU's?

yes. time spent 5-6 months

Try
4e and 5d too
How much GPUs you have ?

[ee1234ee]

I see from the introduction that the 4090 graphics card has 16384 CUDA cores and can perform calculations of 50 trillion per second. Why can't VanitySearch achieve its maximum performance? If VanitySearch can reach speeds of hundreds of thousands of seconds per second on the 4090 graphics card, how great would it be?
Can anyone optimize the program?

[brainless]

I see from the introduction that the 4090 graphics card has 16384 CUDA cores and can perform calculations of 50 trillion per second. Why can't VanitySearch achieve its maximum performance? If VanitySearch can reach speeds of hundreds of thousands of seconds per second on the 4090 graphics card, how great would it be?
Can anyone optimize the program?

50 trillion, where is source of this information ?

[SWATMiner]

Chat GPT 5 with its massive computing power behind it is incapable of searching for WIF for puzzles 71-74.

Question to Chat GPT 5: For bitcoin puzzle numbers from 71 to 74 the first 31 characters of the wallet import format is the same. Only the last part keeps changing. Puzzle no. 70 has been solved and puzzle number 75 has been solved and WIF for these puzzles are available. Using these details is it possible to search for the WIF for puzzle number 71, 72, 73 and 74?

Chat GPT's Answer:
Context on the bitcoin puzzle series
Nature of the puzzles: The “~1000 BTC Bitcoin Challenge” keys were drawn from narrowed private‑key ranges to demonstrate search difficulty; there’s no exploitable pattern beyond the constrained keyspace.

Current status: #70 and #75 have been solved; #71–#74 remain unsolved. Puzzle #71 targets the address 1PWo3JeB9jrGwfHDNpdGK54CRas7fsVzXU2.

Keyspace for #71: The widely cited range is hex 400000000000000000 to 7fffffffffffffffff (i.e., 70–71‑bit interval), making #71 relatively “narrow” compared to higher puzzles, but still astronomically large2.

What the shared WIF prefix actually means
WIF structure: WIF is Base58Check over a payload: 0x80 + private key + optional 0x01 (compressed), plus a 4‑byte checksum. Small changes in the private key produce large changes in the Base58 output.

Shared first 31 characters: If the first 31 WIF characters match across #71–#74, that indicates a shared prefix in the underlying payload bytes (e.g., same version byte and very similar high‑order key bits), but it does not leak the remaining key material in any way that’s useful for deriving the full private key. The checksum also prevents “nearby” guesses from clustering.

Implication: A common WIF prefix does not reduce the brute‑force search space. You still need to find the exact private key inside the specified range and then encode it to WIF.

Is it possible to search for the WIF for #71–#74?
Yes, by brute force over the defined ranges: The only viable approach is to iterate private keys in the puzzle’s keyspace, derive the corresponding address, and check for funds. If found, encode the private key to WIF. Tools and methods referenced by the community include CPU/GPU scanners over the published keyspace ranges.

No, via pattern inference from #70 and #75: The fact that adjacent puzzles (#70 and #75) are solved and their WIFs known does not provide a shortcut to #71–#74. The puzzle creator explicitly stated “there is no pattern,” and keys are consecutive from a deterministic wallet with masking; you still must search the interval.

Practical approach if you want to attempt #71–#74
Define the exact keyspace per puzzle: For #71, use the published range 400000000000000000:7fffffffffffffffff; similar bounded ranges apply to adjacent puzzles, but confirm specifics for each index2.

Use a range scanner: Point a scanner to the keyspace and verify addresses, e.g., community references show command examples with BitCrack or KeyHunt for #71’s range and target address 1PWo3JeB9jrGwfHDNpdGK54CRas7fsVzXU.

Expect extreme difficulty: Even “narrow” ranges at 71–74 bits are enormous. Public “pools” scanning #71 show minute progress relative to the range size, underscoring the scale.

Bottom line
You can search, but only by exhaustive keyspace scanning. Shared WIF prefixes and solved neighbors (#70, #75) don’t grant leverage; they don’t shrink the search. The only path is brute force over the published ranges, which remain prohibitively large in practice2.

[SWATMiner]

I tried something similar with deepseek. It threw out the four private keys in just a couple of seconds. I was elated at first but then when I tried to sweep using the private keys, I discovered that all these seemingly Artificially Intelligent systems are not so intelligent after all!

[kTimesG]

I discovered that all these seemingly Artificially Intelligent systems are not so intelligent after all!

AI does not have computing power, computers have computing power. A dumb ass trained monkey that can teach you how to split a banana doesn't know how to count 1 to 10, let alone break cryptography.

But they're very good at lowering the overall non-artificial intelligence levels though, if you know what I mean.

[Nodemath]

Still I have bid option to the 99.9 percent logic or 99 percent Logic
Start from 1 Btc

Unable to compute that remaining Logic achieving me huge trillion difference


Trust the math's . It never disappoint u

[skedarve]

Still I have bid option to the 99.9 percent logic or 99 percent Logic
Start from 1 Btc

Unable to compute that remaining Logic achieving me huge trillion difference


Trust the math's . It never disappoint u

im still waiting for my BTC  BTC

[brainless]

Still I have bid option to the 99.9 percent logic or 99 percent Logic
Start from 1 Btc

Unable to compute that remaining Logic achieving me huge trillion difference


Trust the math's . It never disappoint u

im still waiting for my BTC   BTC

135 to 155 with me, just waiting for 160 to solve, then I will transfer all together , even someone can solve and get before me
Btw long time ago reward not upgraded
And one have news for this updation like 13.5 to 135
Pls inform us, if near future aspect then I will wait for reward upgraded

[vhh]

Brainless, does it hurt to be you?

[brainless]

Still I have bid option to the 99.9 percent logic or 99 percent Logic
Start from 1 Btc

Unable to compute that remaining Logic achieving me huge trillion difference


Trust the math's . It never disappoint u

im still waiting for my BTC  BTC

I am also still waiting my btc too from BTC-E.com 1 acct 1733 btc 2nd acct 87 btc3rd acct 40+, any news related to victim of btc-e, as heared owner backed to Russia in Feb 2025, but still no announcement.....

[Garys27]

Anyone willing to share their best hash160 prefix matches?

I have
f6f5431d25bbed54360606f002ab373b8ac542c8 KwDiBf89QgGbjEhKnhXJuH7LrciVrZi3rBinyJ7SyJ1aHWsGeWtv
f6f5431d25bbef7104e408a6d529af82ed8db72e KwDiBf89QgGbjEhKnhXJuH7LrciVrZi3rA4qGp3FwkWn5ioNAdA2

[k2laci]

[2025-10-11-14:40:33] [Info] [v1.4.1.0] [vanitysearch] starting... Puzzle: [71]
[2025-10-11-14:40:33] [Info] Target address: 1PWo3JeB9jrGwfHDNpdGK54CRas7fsVzXU
[2025-10-11-14:40:33] [Info] Test mode is passive.
[2025-10-11-14:40:33] [Info] Custom range: [none]
[2025-10-11-14:40:33] [Info] API share: False / Telegram share: False
[2025-10-11-14:40:33] [Info] Untrusted computer: False
[2025-10-11-14:40:33] [Info] Progress: Visit the [btcpuzzle.info] for statistics.
[2025-10-11-14:40:33] [Info] Worker name: worker7539
------------------------------------------------------------------------------------------------------------------------
[2025-10-11-14:40:40] [Info] App [vanitysearch] exited with [-1073741676] code.
                                                                                 


What is the problem ?