Bitcoin puzzle transaction ~32 BTC prize to who solves it

[kTimesG]

To be able to use keyhunt's bsgs mode to scan and solve puzzle 135 as fast as possible, how many CPU threads and RAM are needed at the same time? How long does it take to scan the range 4000000000000000000000000000000000:4ffffffffffffffffffffffffffffffffffff?

Puzzle 135 range size: 2¹³⁴

BSGS requirements:

1. Fast memory baby table: sqrt(N) items = 2⁶⁷ items
Item size: 256 bits in full; let's assume we only need the first 67 bits and ignore hash collision overhead

Total memory required: 2⁶⁷ * 67 bits = somewhere between 2⁷³ to 2⁷⁴ bits

The total amount of data stored on Earth in 2018 was 33 zettabytes. That is, 2⁷⁸ bits.

Estimates for 2025 are around 175 zettabytes. That s, all of the hard drives that exist on Earth have some total capacity of around 2⁸⁰ bits.

The amount of RAM is less than a fraction of all that (if you don't believe this, check any PC: what's the ratio between RAM and storage capacity?)

2. How many threads?

EC operations required at most: sqrt(N) for baby steps + sqrt(N) for giant steps = 2 * sqrt(N) steps

That is, 2⁶⁸ elliptic curve group operations.

A high-end CPU can do around 15 to 20 Mo/s per thread. However, we also need to check the table after every giant step. I will ignore this and assume it is a no-op (it's not).

Total threads needed to solve in one second: 2⁶⁸ / 20.000.000 = 14,757,395,258,968

In summary:

RAM: 1 to 2 zettabytes
CPU threads: 14.8 trillion (for 1 second of total work)

[Akito S. M. Hosana]

The numbers involved are so large and abnormal that half the people in this thread believe it’s a conspiracy whenever someone claims BTC from a puzzle. Some even think the creator is taking the prize for themselves

[mitkopasa]

puzzle 66 was found by a newbie and stolen by bots.
Now Mara has been discovered. But there are not many people who know how to use it. So helping people
If anyone knows how to use Mara. A detailed explanation would be helpful. Maybe someone lucky will find it, don't let the bots lose it.

1- First, you perform the transaction here. You take raw TX out.
https://bitcoin.stackexchange.com/questions/103452/how-to-create-a-signed-transaction-offline-using-electrum

2- Go to slipstream.mara.com and paste it, and follow what is given to you.

It's that simple.

Thanks for the info. but according to the link you gave, won't the publickey be exposed when I sign raw tx in my electrum? in this way, aren't we actually broadcasting the transaction? Also, I didn't see such a step in the new version electrum. the question I asked may seem simple and funny to you, but I am trying to understand how people like me who don't know much can achieve this.

[cctv5go]

puzzle 66 was found by a newbie and stolen by bots.
Now Mara has been discovered. But there are not many people who know how to use it. So helping people
If anyone knows how to use Mara. A detailed explanation would be helpful. Maybe someone lucky will find it, don't let the bots lose it.

1- First, you perform the transaction here. You take raw TX out.
https://bitcoin.stackexchange.com/questions/103452/how-to-create-a-signed-transaction-offline-using-electrum

2- Go to slipstream.mara.com and paste it, and follow what is given to you.

It's that simple.

Thanks for the info. but according to the link you gave, won't the publickey be exposed when I sign raw tx in my electrum? in this way, aren't we actually broadcasting the transaction? Also, I didn't see such a step in the new version electrum. the question I asked may seem simple and funny to you, but I am trying to understand how people like me who don't know much can achieve this.

You can easily create original transaction hexadecimal Tx using Bitcoin Core Wallet or Electum Wallet

[nomachine]

The numbers involved are so large and abnormal that half the people in this thread believe it’s a conspiracy whenever someone claims BTC from a puzzle. Some even think the creator is taking the prize for themselves

You're not thinking like that, are you? Or do you believe that using a transaction on slipstream.mara.com justifies the entire process? Can some activity on Twitter really cover everything? Social media can sometimes amplify misinformation or speculation, so it’s crucial to cross-check information with reliable sources or on-chain data. It’s understandable why some people might think that way, especially when dealing with large sums of money or complex systems like Bitcoin. Sometimes, it’s better to go fishing than to chase conspiracies...  

[WanderingPhilospher]

puzzle 66 was found by a newbie and stolen by bots.
Now Mara has been discovered. But there are not many people who know how to use it. So helping people
If anyone knows how to use Mara. A detailed explanation would be helpful. Maybe someone lucky will find it, don't let the bots lose it.

1- First, you perform the transaction here. You take raw TX out.
https://bitcoin.stackexchange.com/questions/103452/how-to-create-a-signed-transaction-offline-using-electrum

2- Go to slipstream.mara.com and paste it, and follow what is given to you.

It's that simple.

Thanks for the info. but according to the link you gave, won't the publickey be exposed when I sign raw tx in my electrum? in this way, aren't we actually broadcasting the transaction? Also, I didn't see such a step in the new version electrum. the question I asked may seem simple and funny to you, but I am trying to understand how people like me who don't know much can achieve this.

Did you read what I wrote, months back?

https://bitcointalk.org/index.php?topic=1306983.msg64379149#msg64379149

[bibilgin]

Thanks for the info. but according to the link you gave, won't the publickey be exposed when I sign raw tx in my electrum? in this way, aren't we actually broadcasting the transaction? Also, I didn't see such a step in the new version electrum. the question I asked may seem simple and funny to you, but I am trying to understand how people like me who don't know much can achieve this.

If you sign, the publickey is not released. If you say PUBLISH, the transfer starts and the PUBLICKEY is released.

In the new versions, PUBLISH does not come out without SIGNING. I know that.

[Bram24732]

The numbers involved are so large and abnormal that half the people in this thread believe it’s a conspiracy whenever someone claims BTC from a puzzle. Some even think the creator is taking the prize for themselves

Well I can tell you I’m definitely not the creator

[hoanghuy2912]

To be able to use keyhunt's bsgs mode to scan and solve puzzle 135 as fast as possible, how many CPU threads and RAM are needed at the same time? How long does it take to scan the range 4000000000000000000000000000000000:4ffffffffffffffffffffffffffffffffffff?

Puzzle 135 range size: 2¹³⁴

BSGS requirements:

1. Fast memory baby table: sqrt(N) items = 2⁶⁷ items
Item size: 256 bits in full; let's assume we only need the first 67 bits and ignore hash collision overhead

Total memory required: 2⁶⁷ * 67 bits = somewhere between 2⁷³ to 2⁷⁴ bits

The total amount of data stored on Earth in 2018 was 33 zettabytes. That is, 2⁷⁸ bits.

Estimates for 2025 are around 175 zettabytes. That s, all of the hard drives that exist on Earth have some total capacity of around 2⁸⁰ bits.

The amount of RAM is less than a fraction of all that (if you don't believe this, check any PC: what's the ratio between RAM and storage capacity?)

2. How many threads?

EC operations required at most: sqrt(N) for baby steps + sqrt(N) for giant steps = 2 * sqrt(N) steps

That is, 2⁶⁸ elliptic curve group operations.

A high-end CPU can do around 15 to 20 Mo/s per thread. However, we also need to check the table after every giant step. I will ignore this and assume it is a no-op (it's not).

Total threads needed to solve in one second: 2⁶⁸ / 20.000.000 = 14,757,395,258,968

In summary:

RAM: 1 to 2 zettabytes
CPU threads: 14.8 trillion (for 1 second of total work)

really feel helpless, i don't think that puzzles like this have anyone to solve or there is a conspiracy behind it

[kTimesG]

RAM: 1 to 2 zettabytes
CPU threads: 14.8 trillion (for 1 second of total work)

really feel helpless, i don't think that puzzles like this have anyone to solve or there is a conspiracy behind it

BSGS is useless above 80 or so bits, and it's also impossible to scale due to the fast (machine local) table lookup requirement. Some people think that BSGS can scale after some dark magic nonsense, and that Kangaroo is the one that doesn't scale going up, but those people live in a different reality. Meanwhile all puzzles above 70 bits up to 130 bits were solved with Kangaroo-based implementations.

There is at least somebody who most likely is already far ahead in solving 135 by other means, his name is RetiredCoder. He's probably the reason why you may find it hard to find a cheap RTX 4090 instance for renting. The only reason no one else doesn't have much chances, despite having the software to do it, is because they aren't millionaires.

[TKDomino]

Question for someone smarter than me. I don't understand, I'm running KeyHunt-Cuda but the output shows 661.92 Mk/s but for the T: 15,032,385,536. only 15 million. can someone explain it because the original key hunt was display correctly. maybe its just me not understanding what the T stands for. I though it was threads processed which would be a single key processed. correct me if I'm wrong please.

PS C:\KeyHunt-Cuda-main> .\KeyHunt-Cuda.exe -t 0 -g --gpui 0 --gpux 512,512 -m address --coin BTC --range 730fc235c00000000:730fc235fffffffff 1BY8GQbnueYofwSuFAT3USAhGjPrkxDdW9

KeyHunt-Cuda v1.07

COMP MODE    : COMPRESSED
COIN TYPE    : BITCOIN
SEARCH MODE  : Single Address
DEVICE       : GPU
CPU THREAD   : 0
GPU IDS      : 0
GPU GRIDSIZE : 512x512
SSE          : YES
RKEY         : 0 Mkeys
MAX FOUND    : 65536
BTC ADDRESS  : 1BY8GQbnueYofwSuFAT3USAhGjPrkxDdW9
OUTPUT FILE  : Found.txt

Start Time   : Sat Feb 22 13:25:47 2025
Global start : 730FC235C00000000 (67 bit)
Global end   : 730FC235FFFFFFFFF (67 bit)
Global range : 3FFFFFFFF (34 bit)

GPU          : GPU #0 Tesla T4 (40x64 cores) Grid(512x512)

[00:00:20] [CPU+GPU: 661.98 Mk/s] [GPU: 661.98 Mk/s] [C: 78.125000 %] [R: 0] [T: 13,421,772,800 (34 bit)] [F: 0]
=================================================================================
PubAddress: 1BY8GQbnueYofwSuFAT3USAhGjPrkxDdW9
Priv (WIF): p2pkh:KwDiBf89QgGbjEhKnhXJuH7LrciVrZi3qbP2K5cm35XKMND1X1KW
Priv (HEX): 730FC235C1942C1AE
PubK (HEX): 0212209F5EC514A1580A2937BD833979D933199FC230E204C6CDC58872B7D46F75
=================================================================================
[00:00:22] [CPU+GPU: 661.92 Mk/s] [GPU: 661.92 Mk/s] [C: 87.500000 %] [R: 0] [T: 15,032,385,536 (34 bit)] [F: 1]

[madogss]

Question for someone smarter than me. I don't understand, I'm running KeyHunt-Cuda but the output shows 661.92 Mk/s but for the T: 15,032,385,536. only 15 million. can someone explain it because the original key hunt was display correctly. maybe its just me not understanding what the T stands for. I though it was threads processed which would be a single key processed. correct me if I'm wrong please.

PS C:\KeyHunt-Cuda-main> .\KeyHunt-Cuda.exe -t 0 -g --gpui 0 --gpux 512,512 -m address --coin BTC --range 730fc235c00000000:730fc235fffffffff 1BY8GQbnueYofwSuFAT3USAhGjPrkxDdW9

KeyHunt-Cuda v1.07

COMP MODE    : COMPRESSED
COIN TYPE    : BITCOIN
SEARCH MODE  : Single Address
DEVICE       : GPU
CPU THREAD   : 0
GPU IDS      : 0
GPU GRIDSIZE : 512x512
SSE          : YES
RKEY         : 0 Mkeys
MAX FOUND    : 65536
BTC ADDRESS  : 1BY8GQbnueYofwSuFAT3USAhGjPrkxDdW9
OUTPUT FILE  : Found.txt

Start Time   : Sat Feb 22 13:25:47 2025
Global start : 730FC235C00000000 (67 bit)
Global end   : 730FC235FFFFFFFFF (67 bit)
Global range : 3FFFFFFFF (34 bit)

GPU          : GPU #0 Tesla T4 (40x64 cores) Grid(512x512)

[00:00:20] [CPU+GPU: 661.98 Mk/s] [GPU: 661.98 Mk/s] [C: 78.125000 %] [R: 0] [T: 13,421,772,800 (34 bit)] [F: 0]
=================================================================================
PubAddress: 1BY8GQbnueYofwSuFAT3USAhGjPrkxDdW9
Priv (WIF): p2pkh:KwDiBf89QgGbjEhKnhXJuH7LrciVrZi3qbP2K5cm35XKMND1X1KW
Priv (HEX): 730FC235C1942C1AE
PubK (HEX): 0212209F5EC514A1580A2937BD833979D933199FC230E204C6CDC58872B7D46F75
=================================================================================
[00:00:22] [CPU+GPU: 661.92 Mk/s] [GPU: 661.92 Mk/s] [C: 87.500000 %] [R: 0] [T: 15,032,385,536 (34 bit)] [F: 1]

That's displaying correctly, that is 15 billon not million multiply your Mk/s by the time so 661920000 * 22 = 14.56 billion this number is not exact because the speed fluctuates.

The T is for total keys checked.

[TKDomino]

Question for someone smarter than me. I don't understand, I'm running KeyHunt-Cuda but the output shows 661.92 Mk/s but for the T: 15,032,385,536. only 15 million. can someone explain it because the original key hunt was display correctly. maybe its just me not understanding what the T stands for. I though it was threads processed which would be a single key processed. correct me if I'm wrong please.

PS C:\KeyHunt-Cuda-main> .\KeyHunt-Cuda.exe -t 0 -g --gpui 0 --gpux 512,512 -m address --coin BTC --range 730fc235c00000000:730fc235fffffffff 1BY8GQbnueYofwSuFAT3USAhGjPrkxDdW9

KeyHunt-Cuda v1.07

COMP MODE    : COMPRESSED
COIN TYPE    : BITCOIN
SEARCH MODE  : Single Address
DEVICE       : GPU
CPU THREAD   : 0
GPU IDS      : 0
GPU GRIDSIZE : 512x512
SSE          : YES
RKEY         : 0 Mkeys
MAX FOUND    : 65536
BTC ADDRESS  : 1BY8GQbnueYofwSuFAT3USAhGjPrkxDdW9
OUTPUT FILE  : Found.txt

Start Time   : Sat Feb 22 13:25:47 2025
Global start : 730FC235C00000000 (67 bit)
Global end   : 730FC235FFFFFFFFF (67 bit)
Global range : 3FFFFFFFF (34 bit)

GPU          : GPU #0 Tesla T4 (40x64 cores) Grid(512x512)

[00:00:20] [CPU+GPU: 661.98 Mk/s] [GPU: 661.98 Mk/s] [C: 78.125000 %] [R: 0] [T: 13,421,772,800 (34 bit)] [F: 0]
=================================================================================
PubAddress: 1BY8GQbnueYofwSuFAT3USAhGjPrkxDdW9
Priv (WIF): p2pkh:KwDiBf89QgGbjEhKnhXJuH7LrciVrZi3qbP2K5cm35XKMND1X1KW
Priv (HEX): 730FC235C1942C1AE
PubK (HEX): 0212209F5EC514A1580A2937BD833979D933199FC230E204C6CDC58872B7D46F75
=================================================================================
[00:00:22] [CPU+GPU: 661.92 Mk/s] [GPU: 661.92 Mk/s] [C: 87.500000 %] [R: 0] [T: 15,032,385,536 (34 bit)] [F: 1]

That's displaying correctly, that is 15 billon not million multiply your Mk/s by the time so 661920000 * 22 = 14.56 billion this number is not exact because the speed fluctuates.

The T is for total keys checked.

So, if I'm understanding it the last 3 places are cut off or will they continue to be cut off as the number grows. The reason why I'm asking is to keep track of my place. What would the correct syntax to output it to a file and would the file have it cut off or would it be correct in the file?

[madogss]

So, if I'm understanding it the last 3 places are cut off or will they continue to be cut off as the number grows. The reason why I'm asking is to keep track of my place. What would the correct syntax to output it to a file and would the file have it cut off or would it be correct in the file?

No the last 3 places are not cut off and the number continues to grow without hiding any numbers.

It is hard to keep track of your place because keyhunt-cuda works by dividing the global range by your grid so each thread is assigned a small chunk and each is incremented.

example if your grid is 512,512 and your global range is 3FFFFFFFF (34 bit) then that means you have 512 * 512 = 262,144 threads
so with 3FFFFFFFF(17179869183) / 262,144 = 65,535.999996185302734375 then we round to 65,536
this means each thread searches a small range of 65,536 keys.

[Bram24732]

Hey, BTC67 winner here. What's the best way for me to contact RetiredCoder ?

IFUCyxSGddzbuAxOmhrLBQ+4Q606tFU81wRu8wWg30VxHNNDcKGlcHDJH4aRTnxFE6W8Xc6VPtVQxw+DSadYKlk=

[mitkopasa]

puzzle 66 was found by a newbie and stolen by bots.
Now Mara has been discovered. But there are not many people who know how to use it. So helping people
If anyone knows how to use Mara. A detailed explanation would be helpful. Maybe someone lucky will find it, don't let the bots lose it.

1- First, you perform the transaction here. You take raw TX out.
https://bitcoin.stackexchange.com/questions/103452/how-to-create-a-signed-transaction-offline-using-electrum

2- Go to slipstream.mara.com and paste it, and follow what is given to you.

It's that simple.

Thanks for the info. but according to the link you gave, won't the publickey be exposed when I sign raw tx in my electrum? in this way, aren't we actually broadcasting the transaction? Also, I didn't see such a step in the new version electrum. the question I asked may seem simple and funny to you, but I am trying to understand how people like me who don't know much can achieve this.

Did you read what I wrote, months back?

https://bitcointalk.org/index.php?topic=1306983.msg64379149#msg64379149

thank you WanderingPhilospher and bibilgin. I think I will test it first with a small amount of value.

[Baskentliia]

puzzle 66 was found by a newbie and stolen by bots.
Now Mara has been discovered. But there are not many people who know how to use it. So helping people
If anyone knows how to use Mara. A detailed explanation would be helpful. Maybe someone lucky will find it, don't let the bots lose it.

1- First, you perform the transaction here. You take raw TX out.
https://bitcoin.stackexchange.com/questions/103452/how-to-create-a-signed-transaction-offline-using-electrum

2- Go to slipstream.mara.com and paste it, and follow what is given to you.

It's that simple.

Thanks for the info. but according to the link you gave, won't the publickey be exposed when I sign raw tx in my electrum? in this way, aren't we actually broadcasting the transaction? Also, I didn't see such a step in the new version electrum. the question I asked may seem simple and funny to you, but I am trying to understand how people like me who don't know much can achieve this.

Did you read what I wrote, months back?

https://bitcointalk.org/index.php?topic=1306983.msg64379149#msg64379149

thank you WanderingPhilospher and bibilgin. I think I will test it first with a small amount of value.

Is there any chance you can take screenshots and videos? It will be very useful for friends who don't know.
After shooting video, you move the assets to a secure address and then publish the video and images

[Niekko]

To be able to use keyhunt's bsgs mode to scan and solve puzzle 135 as fast as possible, how many CPU threads and RAM are needed at the same time? How long does it take to scan the range 4000000000000000000000000000000000:4ffffffffffffffffffffffffffffffffffff?

Puzzle 135 range size: 2¹³⁴

BSGS requirements:

1. Fast memory baby table: sqrt(N) items = 2⁶⁷ items
Item size: 256 bits in full; let's assume we only need the first 67 bits and ignore hash collision overhead

Total memory required: 2⁶⁷ * 67 bits = somewhere between 2⁷³ to 2⁷⁴ bits

The total amount of data stored on Earth in 2018 was 33 zettabytes. That is, 2⁷⁸ bits.

Estimates for 2025 are around 175 zettabytes. That s, all of the hard drives that exist on Earth have some total capacity of around 2⁸⁰ bits.

The amount of RAM is less than a fraction of all that (if you don't believe this, check any PC: what's the ratio between RAM and storage capacity?)

2. How many threads?

EC operations required at most: sqrt(N) for baby steps + sqrt(N) for giant steps = 2 * sqrt(N) steps

That is, 2⁶⁸ elliptic curve group operations.

A high-end CPU can do around 15 to 20 Mo/s per thread. However, we also need to check the table after every giant step. I will ignore this and assume it is a no-op (it's not).

Total threads needed to solve in one second: 2⁶⁸ / 20.000.000 = 14,757,395,258,968

In summary:

RAM: 1 to 2 zettabytes
CPU threads: 14.8 trillion (for 1 second of total work)

Mathematics is like sex... if you go too fast, you miss the best part. 

[karrask]

Deepseek how do you know bc1qgp48hjxp9uctzysq458dtlhk7ewtf9k4xpjpjj is the creator, their reason for sending 184USD TO #66 is not clear, but you are assuming it is a clue to #67?

Also if it was a clue why ignore the zeros, 0.00189717 = 0.007C553B1ADE27BE0A11   and 0.00010392  =  0.0006CF7D005BC5789A9B

I think you stretching cause as far as I know nobody ever correctly guessed #66 started with 283 , how could they guess #67 but not #66

https://www.talkimg.com/images/2025/02/17/qMGlW.png

I was about 8 hours short of opening 66, I was rummaging through this range that day. It's a shame.
I started with the logarithm 19.666 and didn't get there. Which was found through triangles in AutoCAD.
Sry i use translater.

Hi. Friend, let's talk without an interpreter. I have a couple of questions.

[Stanislav01]

Hey, BTC67 winner here. What's the best way for me to contact RetiredCoder ?

IFUCyxSGddzbuAxOmhrLBQ+4Q606tFU81wRu8wWg30VxHNNDcKGlcHDJH4aRTnxFE6W8Xc6VPtVQxw+DSadYKlk=

Hello! Congratulations!