BTC PUZZLE

[the crack]

I am working on a modified version of kangaroo, want to know how much time will it take to solve puzzle 53 (i know its already solved, i am testing on it) on mac mini M4 with normal kangaroo script?

[nc50lc]

Should it have to be specific to that expensive piece of hardware?

In case no one with 'mac mini M4' replies (so far, no one after 34hours), you can ask for benchmarks from others and research the specs for comparison.
Or specify other machines you have, one with easy to access CPU.

I'll add my best result with 8th-gen i7:

Code:

Kangaroo v2.2
Start:EFAE164CB9E3D
Stop :1FFFFFFFFFFFFF
Keys :1
Number of CPU thread: 12
Range width: 2^53
Jump Avg distance: 2^25.96
Number of kangaroos: 2^13.58
Suggested DP: 10
Expected operations: 2^27.61
Expected RAM: 19.7MB
DP size: 10 [0xFFC0000000000000]
SolveKeyCPU Thread 2: 1024 kangaroos
SolveKeyCPU Thread 8: 1024 kangaroos
SolveKeyCPU Thread 3: 1024 kangaroos
SolveKeyCPU Thread 5: 1024 kangaroos
SolveKeyCPU Thread 4: 1024 kangaroos
SolveKeyCPU Thread 11: 1024 kangaroos
SolveKeyCPU Thread 6: 1024 kangaroos
SolveKeyCPU Thread 0: 1024 kangaroos
SolveKeyCPU Thread 9: 1024 kangaroos
SolveKeyCPU Thread 7: 1024 kangaroos
SolveKeyCPU Thread 1: 1024 kangaroos
SolveKeyCPU Thread 10: 1024 kangaroos
[52.27 MK/s][GPU 0.00 MK/s][Count 2^26.97][Dead 1][02s (Avg 03s)][5.9/20.3MB]
Done: Total time 03s

CPU: Intel i7 8700 = 03sec (averages to 4 seconds) | Yours should be faster using the same software.
Using: https://github.com/JeanLucPons/Kangaroo (does this count as "normal" kangaroo script?)

[NotATether]

An Apple M4? Nobody has a clue. Most likely because nobody has written any solvers for Apple Silicon's ARM64 yet. However, I am sure that it will be several times faster than a recent Core i7 (or whatever the hell Intel calls them now)

The M4 (and by extension, the M1, M2 and M3) have a GPU integrated into the die, so while I've assumed that the general-purpose processor will outperform an Intel CPU, leveraging the GPU as well could bring it close to Nvidia-levels of speed. Or maybe exceed it. Like I said, nobody knows yet.

[kTimesG]

An Apple M4? Nobody has a clue. Most likely because nobody has written any solvers for Apple Silicon's ARM64 yet. However, I am sure that it will be several times faster than a recent Core i7 (or whatever the hell Intel calls them now)

The M4 (and by extension, the M1, M2 and M3) have a GPU integrated into the die, so while I've assumed that the general-purpose processor will outperform an Intel CPU, leveraging the GPU as well could bring it close to Nvidia-levels of speed. Or maybe exceed it. Like I said, nobody knows yet.

On a M1 (2020!), using libsecp256k1 mods to do kangaroo jumps (e.g. optimized batched point additions) runs at 11 Mo/s per thread. So around 60-70 Mo/s.

I think the GPU is shared with CPU resources, but it has a really low teraflops, like just a fraction of a laptop RTX card. The benefits however are that the performance/watt is so damn good that you never hear the fan even when the chip runs at 100% (while it's a total nightmare to be in the same room with a RTX laptop that crunches numbers). But of course I would never ever recommend to use any kind of Mac as a non-stop kangaroo node, only for development in a zen environment.

[coolfree]

An Apple M4? Nobody has a clue. Most likely because nobody has written any solvers for Apple Silicon's ARM64 yet. However, I am sure that it will be several times faster than a recent Core i7 (or whatever the hell Intel calls them now)

The M4 (and by extension, the M1, M2 and M3) have a GPU integrated into the die, so while I've assumed that the general-purpose processor will outperform an Intel CPU, leveraging the GPU as well could bring it close to Nvidia-levels of speed. Or maybe exceed it. Like I said, nobody knows yet.

Apple Silicon is going to totally change everything --- not just bitcoins.

Stock up. An M2x4 should do fine.

[Gourav77]

I have a low end cpu I have i5 🙃 and I wanna try to solve one of the unsolved puzzle but I guess I'm out of this because I have a low end device . Can anyone give me suggestions if there is any possibility 🙃 I am a student currently in 3 year of computer science...

[kTimesG]

I have a low end cpu I have i5 🙃 and I wanna try to solve one of the unsolved puzzle but I guess I'm out of this because I have a low end device . Can anyone give me suggestions if there is any possibility 🙃 I am a student currently in 3 year of computer science...

Possibile: yes.
Probability to solve an unsolved puzzle with a single i5 CPU in your lifetime: there are higher chances you notice an alien ship invading while you are reading this, but an unexpected sun burst obliterates it just seconds before it fires its quantum weapon. Then a giant wormhole caused by some secret project absorbs all of the the plasma burst fireworks, and everyone gets free electricity for a few centuries out of it. Oh, and all this should happen just after scientists found a cure for cancer.

67 was solved by using an immense amount of GPU power (thousands) running for several months, what do you expect even with the fastest of CPUs? You'd need a bunch of millions of them to stand any chance.

[nc50lc]

I have a low end cpu I have i5 🙃 and I wanna try to solve one of the unsolved puzzle but I guess I'm out of this because I have a low end device .

True, let's say that your i5 can do 20MK/s with Kangaroo (ECDLP Solver) and you're targeting puzzle #135 (pubKey exposed) which has about 67.5-bit security against it,
Your chances would be: 1/10,435,054,260,266 or 0.00000000000009583%

If you're brute-forcing Puzzle #68 instead at the same rate,
Your chances would be: 1/14,757,395,258,967 or 0.00000000000006776%

Can anyone give me suggestions if there is any possibility 🙃 I am a student currently in 3 year of computer science...

Since you have a background in Comp-Sci, I think you can already tell that it's close-to-zero with that hardware based from the numbers above.