brute-forcing public keys at amazing speed 2.2 PH/s on CPU [malware warning]

[COBRAS]

What if use many publick key from 1 adress ? Maby jrivate key will be 1 and private key search for many public adreses will get more trustwothy result ?

Can someone test this idea ?

[MrFreeDragon]

Yes I remarked that this first key of this set often generate dead kangaroos (even in 64bit range)
The fact that the startrange is not zero should not impact, it is just a translation. It is equivalent to what MrFreeDragon said.
-snip-

Finally I could solve the 80bit key from the 16key example for 3h14min. But I made the adjustment I wrote before.

Changed the range in the following way:

Code:

Real ranges:
Start:49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB4800000000000000000000
Stop :49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48FFFFFFFFFFFFFFFFFFFF

Changed to (deducted the Start Range value):
Start:0
Stop :FFFFFFFFFFFFFFFFFFFF

And also made the corrsponding changes to the public keys: deducted the Point 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB4800000000000000000000*G from every Public key, so:

Code:

The initial pubic keys:
0459A3BFDAD718C9D3FAC7C187F1139F0815AC5D923910D516E186AFDA28B221DC994327554CED887AAE5D211A2407CDD025CFC3779ECB9C9D7F2F1A1DDF3E9FF8
04A50FBBB20757CC0E9C41C49DD9DF261646EE7936272F3F68C740C9DA50D42BCD3E48440249D6BC78BC928AA52B1921E9690EBA823CBC7F3AF54B3707E6A73F34
0404A49211C0FE07C9F7C94695996F8826E09545375A3CF9677F2D780A3EB70DE3BD05357CAF8340CB041B1D46C5BB6B88CD9859A083B0804EF63D498B29D31DD1
040B39E3F26AF294502A5BE708BB87AEDD9F895868011E60C1D2ABFCA202CD7A4D1D18283AF49556CF33E1EA71A16B2D0E31EE7179D88BE7F6AA0A7C5498E5D97F
04837A31977A73A630C436E680915934A58B8C76EB9B57A42C3C717689BE8C0493E46726DE04352832790FD1C99D9DDC2EE8A96E50CAD4DCC3AF1BFB82D51F2494

were replaced by these ones:
046cf3ec949c28918ed0fdc48980b3339343ceb822acd9b7b3a0d16d9606aeb139507c48c74b81fde39c415117f914fd850e8a64bde218e99b6d74b2012977df56
04948282bfa4cc7aaa1722f0f72ce9b3708061960084ba188a2c66713c46070f45c9a997916266123aacbc5fbc76e1a0a133ab92410466f19f92e67ab1716f0db4
0491a7b757335e7026e83c42b868dd6c4a8f379e7ef059de9a1fc1b7e4e770749ae1ed4034f6d24d0da3d7ce10ce3c543dade4b812f118e0d0897ca2d42aa1e3d7
0420c453ffcc3840c90c33f695bb1d573d2860c2203c75808ae45a168aa17995af952bea588e1362d9cefe0977a96ea79f554d5809291c20339b3427b55c48859f
0432b9ca00d710738587a4ac0f7aad49516e0ca04ea1a19e9548ae29f00675df84963ba73af8fa4e3bb969c438150ae9f1fc6f3957b3b8be9aab78f19ff7caf20a

And the result for the 1st key was found for 3:14 hours:

Code:

Start:0
Stop :FFFFFFFFFFFFFFFFFFFF
Keys :5
Number of CPU thread: 0
Range width: 2^80
Number of random walk: 2^20.81 (Max DP=17)
DP size: 17 [0xffff800000000000]
GPU: GPU #0 GeForce GTX 1080 Ti (28x128 cores) Grid(56x256) (177.0 MB used)
SolveKeyGPU Thread GPU#0: creating kangaroos...
SolveKeyGPU Thread GPU#0: 2^20.81 kangaroos
[436.47 MKey/s][GPU 436.47 MKey/s][Count 2^42.02][Dead 8][03:14:17][2606.2MB]  
Key# 0 Pub:  0x026CF3EC949C28918ED0FDC48980B3339343CEB822ACD9B7B3A0D16D9606AEB139 
       Priv: 0xCB5EBB3EF3883C1866D4

Now we can retrieve the target private key:
Final Priv = 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB4800000000000000000000 + 0xCB5EBB3EF3883C1866D4 = 0x49dccfd96dc5df56487436f5a1b18c4f5d34f65ddb48cb5ebb3ef3883c1866d4

I know that these adsjutments should not affect the time to solve, however in my case without adjustments i could not solve the 1st 80bit from example for 4 hours, with the adjustments the 1st key was solved for 3:14 hours.

If Etar could test that adjusted public keys with the adjusted ranges (with leading zeros) on his 2080ti it would be great. Etar should have the speed 2.5 times faster, so the 1st key should be solved for 1:15-1:30 hours.

[Jean_Luc]

During the night, I solved the key with the same config file as Etar in 2^41.59 iterations (09:06:32) and 1 dead kangaroo.
It seems normal. I will make further tests to see if changing the jumps can improve things a bit...

Code:

C:\C++\Kangaroo\VC_CUDA10\x64\Release>Kangaroo.exe -t 0 -gpu ..\..\in.txt
Kangaroo v1.2
Start:49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB4800000000000000000000
Stop :49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48FFFFFFFFFFFFFFFFFFFF
Keys :1
Number of CPU thread: 0
Range width: 2^80
Number of random walk: 2^18.58 (Max DP=19)
DP size: 19 [0xFFFFE00000000000]
GPU: GPU #0 GeForce GTX 1050 Ti (6x128 cores) Grid(12x256) (45.0 MB used)
SolveKeyGPU Thread GPU#0: creating kangaroos...
SolveKeyGPU Thread GPU#0: 2^18.58 kangaroos in 1376.4ms
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^41.59][Dead 1][09:06:28][485.7MB]
Key# 0 Pub:  0x0259A3BFDAD718C9D3FAC7C187F1139F0815AC5D923910D516E186AFDA28B221DC
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5EBB3EF3883C1866D4

Done: Total time 09:06:32

[PietCoin97]

Code:

Kangaroo.exe -t 0 -gpu -gpuId 0,1,2,3,4,5 in.txt
Kangaroo v1.2
Start:8000000000000000000000001
Stop :FFFFFFFFFFFFFFFFFFFFFFFFE
Keys :2048
Number of CPU thread: 0
Range width: 2^99
Number of random walk: 2^22.49 (Max DP=25)
DP size: 25 [0xFFFFFF8000000000]
GPU: GPU #2 P104-100 (15x128 cores) Grid(30x256) (99.0 MB used)
SolveKeyGPU Thread GPU#2: creating kangaroos...
GPU: GPU #1 P104-100 (15x128 cores) Grid(30x256) (99.0 MB used)
SolveKeyGPU Thread GPU#1: creating kangaroos...
GPU: GPU #0 P104-100 (15x128 cores) Grid(30x256) (99.0 MB used)
SolveKeyGPU Thread GPU#0: creating kangaroos...
GPU: GPU #5 GeForce GTX 1070 (15x128 cores) Grid(30x256) (99.0 MB used)
SolveKeyGPU Thread GPU#5: creating kangaroos...
GPU: GPU #3 GeForce GTX 1070 (15x128 cores) Grid(30x256) (99.0 MB used)
SolveKeyGPU Thread GPU#3: creating kangaroos...
GPU: GPU #4 GeForce GTX 1070 (15x128 cores) Grid(30x256) (99.0 MB used)
SolveKeyGPU Thread GPU#4: creating kangaroos...
SolveKeyGPU Thread GPU#0: 2^19.91 kangaroos in 9236.5ms
SolveKeyGPU Thread GPU#2: 2^19.91 kangaroos in 9508.7ms
SolveKeyGPU Thread GPU#1: 2^19.91 kangaroos in 9569.1ms
SolveKeyGPU Thread GPU#3: 2^19.91 kangaroos in 8538.5ms
SolveKeyGPU Thread GPU#5: 2^19.91 kangaroos in 8782.5ms
SolveKeyGPU Thread GPU#4: 2^19.91 kangaroos in 8431.8ms
[1323.30 MKey/s][GPU 1323.30 MKey/s][Count 2^45.83][Dead 0][14:44:30][147.1MB]

Jean luc does it work also with such an key space ?

Start:8000000000000000000000001
Stop :FFFFFFFFFFFFFFFFFFFFFFFFE

or should i give in all 64 bits ?

and are my setting ok or can i get higher speed ?

[Jean_Luc]

I solved 20 times the following key (64bit range) and there was a strange 24 dead kangaroos event !? It seems that from times to times the algorithm has hard time to solve the key... Just a matter of probability ? My Mersenne twister fails ? I will make further tests...

Key# 0 Pub:  0x0259A3BFDAD718C9D3FAC7C187F1139F0815AC5D923910D516E186AFDA28B221DC
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5EBB3EF3883C1866D4

[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.63][Dead 3][02:12][502.4MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^34.07][Dead 0][02:59][675.8MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^31.43][Dead 0][29s][113.4MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^33.70][Dead 1][02:19][525.5MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^35.20][Dead 24][06:31][1472.8MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^34.44][Dead 7][03:52][872.4MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^33.51][Dead 1][02:02][459.6MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^34.27][Dead 7][03:27][779.1MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^33.68][Dead 3][02:17][516.3MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^33.85][Dead 3][02:34][581.7MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.33][Dead 1][54s][206.9MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.40][Dead 0][57s][216.1MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^33.45][Dead 2][01:57][441.1MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.97][Dead 1][01:24][319.5MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^34.66][Dead 6][04:29][1012.6MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.12][Dead 0][47s][178.9MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^31.66][Dead 0][34s][131.5MB]
[115.23 MKey/s][GPU 115.23 MKey/s][Count 2^33.32][Dead 3][01:47][403.6MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^34.00][Dead 2][02:52][647.1MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.68][Dead 2][02:17][516.5MB]

[Jean_Luc]

Jean luc does it work also with such an key space ?

Start:8000000000000000000000001
Stop :FFFFFFFFFFFFFFFFFFFFFFFFE

or should i give in all 64 bits ?

and are my setting ok or can i get higher speed ?

Yes it should work. Your setup looks good.

[MrFreeDragon]

Tried to use Kangaroo v1.2 on Ubuntu and searched for 2^80 range width. The 1st key was solved for the same tim 3h15min, the 2nd for 4h and the 3d for 4h13min. Here are the results:

Code:

input 5 public keys:
046cf3ec949c28918ed0fdc48980b3339343ceb822acd9b7b3a0d16d9606aeb139507c48c74b81fde39c415117f914fd850e8a64bde218e99b6d74b2012977df56
04948282bfa4cc7aaa1722f0f72ce9b3708061960084ba188a2c66713c46070f45c9a997916266123aacbc5fbc76e1a0a133ab92410466f19f92e67ab1716f0db4
0491a7b757335e7026e83c42b868dd6c4a8f379e7ef059de9a1fc1b7e4e770749ae1ed4034f6d24d0da3d7ce10ce3c543dade4b812f118e0d0897ca2d42aa1e3d7
0420c453ffcc3840c90c33f695bb1d573d2860c2203c75808ae45a168aa17995af952bea588e1362d9cefe0977a96ea79f554d5809291c20339b3427b55c48859f
0432b9ca00d710738587a4ac0f7aad49516e0ca04ea1a19e9548ae29f00675df84963ba73af8fa4e3bb969c438150ae9f1fc6f3957b3b8be9aab78f19ff7caf20a

I used the adjusted keys (deducted the start range from every point), but actually such adjustment does not make sense.

Code:

Kangaroo v1.2
Start:0
Stop :FFFFFFFFFFFFFFFFFFFF
Keys :5
Number of CPU thread: 0
Range width: 2^80
Number of random walk: 2^20.81 (Max DP=17)
DP size: 17 [0xffff800000000000]
GPU: GPU #0 GeForce GTX 1080 Ti (28x128 cores) Grid(56x256) (177.0 MB used)
SolveKeyGPU Thread GPU#0: creating kangaroos...
SolveKeyGPU Thread GPU#0: 2^20.81 kangaroos in 6624.7ms
[443.06 MKey/s][GPU 443.06 MKey/s][Count 2^42.05][Dead 3][03:15:57][2655.9MB]  
Key# 0 Pub:  0x026CF3EC949C28918ED0FDC48980B3339343CEB822ACD9B7B3A0D16D9606AEB139 
       Priv: 0xCB5EBB3EF3883C1866D4 
[457.08 MKey/s][GPU 457.08 MKey/s][Count 2^42.39][Dead 5][04:00:43][3365.8MB]  
Key# 1 Pub:  0x02948282BFA4CC7AAA1722F0F72CE9B3708061960084BA188A2C66713C46070F45 
       Priv: 0xCB5EB5ABC43BEBAD3207 
[456.63 MKey/s][GPU 456.63 MKey/s][Count 2^42.47][Dead 8][04:13:19][3540.3MB]  
Key# 2 Pub:  0x0391A7B757335E7026E83C42B868DD6C4A8F379E7EF059DE9A1FC1B7E4E770749A 
       Priv: 0xCB5E5698AAAB6CAC52B3 

The version 1.2 is faster by 2-3% compared with the version 1.1

I stoped the code and did not wait for the remaining 2 keys. We can see that for 2^80 range it needed 2^42.05, 2^42.39 and 2^42.47 operations, however on average 2^41 was enough. So the total time for every key was 2-3 times longer (!!!) For 2^80 range it makes big difference (1.5 hours vs 3 hours).

But due to the tests made I can confirm that on GTX 1080ti (11Gb) the key in 2^80 range could be found for 3-4 hours, although 1.5 hours on average was expected.

Jean_Luc, are all the wild kangaroos start from the one public key point? I guess that symmetry would halve the total time...

We have to solve P = k.G, we know that k lies in the range [k1,k2];
So, (k2-k) also lies in the range [k1,k2]. How about starting the wild kangaroos not only from k, but from both k and (k2-k)?

[Jean_Luc]

Yes there is a problem with the spreading of wild kangaroo i think.
They should be spread with a -(k2-k1)/2 translation, the keys you solved are close to the upper bound, it explains the factor of 2.
I will try this ASAP.
Many thanks for the tests

[MrFreeDragon]

I solved 20 times the following key (64bit range) and there was a strange 24 dead kangaroos event !? It seems that from times to times the algorithm has hard time to solve the key... Just a matter of probability ? My Mersenne twister fails ? I will make further tests...

Key# 0 Pub:  0x0259A3BFDAD718C9D3FAC7C187F1139F0815AC5D923910D516E186AFDA28B221DC
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5EBB3EF3883C1866D4
-snip-

I tried to do the same (solved 20 times the same key as you). The total result was the same as test with 64bit before: average 1minute per 64bit range key (total 20:52 min for 20 keys).
It seems that my GPU card is more friendly - maximum kangaroo kills was 8 times for the first time.

[300.90 MKey/s][GPU 300.90 MKey/s][Count 2^34.32][Dead 8][01:06][3192.1MB]  
[347.36 MKey/s][GPU 347.36 MKey/s][Count 2^33.94][Dead 7][52s][2462.2MB]  
[456.28 MKey/s][GPU 456.28 MKey/s][Count 2^32.93][Dead 1][26s][1223.2MB]  
[415.46 MKey/s][GPU 415.46 MKey/s][Count 2^33.44][Dead 4][36s][1742.3MB]  
[350.83 MKey/s][GPU 350.83 MKey/s][Count 2^32.23][Dead 0][18s][754.4MB]  
[403.60 MKey/s][GPU 403.60 MKey/s][Count 2^33.52][Dead 2][38s][1837.3MB]  
[299.12 MKey/s][GPU 299.12 MKey/s][Count 2^31.98][Dead 0][16s][635.9MB]  
[458.32 MKey/s][GPU 458.32 MKey/s][Count 2^32.80][Dead 0][26s][1120.4MB]  
[280.27 MKey/s][GPU 280.27 MKey/s][Count 2^34.59][Dead 6][01:28][3846.3MB]  
[203.74 MKey/s][GPU 203.74 MKey/s][Count 2^30.73][Dead 0][10s][270.9MB]  
[296.67 MKey/s][GPU 296.67 MKey/s][Count 2^34.36][Dead 6][01:14][3301.0MB]  
[279.64 MKey/s][GPU 279.64 MKey/s][Count 2^31.67][Dead 0][14s][514.0MB]  
[324.70 MKey/s][GPU 324.70 MKey/s][Count 2^34.13][Dead 4][01:02][2811.4MB]  
[297.73 MKey/s][GPU 297.73 MKey/s][Count 2^34.38][Dead 7][01:14][3339.0MB]  
[426.65 MKey/s][GPU 426.65 MKey/s][Count 2^33.35][Dead 2][34s][1635.3MB]  
[408.92 MKey/s][GPU 408.92 MKey/s][Count 2^32.45][Dead 2][20s][875.8MB]  
[321.43 MKey/s][GPU 321.43 MKey/s][Count 2^34.17][Dead 4][01:02][2876.2MB]  
[250.27 MKey/s][GPU 250.27 MKey/s][Count 2^34.98][Dead 7][02:04][5059.6MB]  
[321.52 MKey/s][GPU 321.52 MKey/s][Count 2^34.16][Dead 2][01:02][2868.8MB]  
[353.13 MKey/s][GPU 353.13 MKey/s][Count 2^33.89][Dead 3][50s][2379.2MB]

Done: Total time 20:52

That "dead" kangaroo case is the cycle running. Kangaroos running in cycles do not find new points, but slow down the algorithm and also might cause it to fail. This was described by Edlyn Teske in Computing discrete logarithms with theparallelized kangaroo method work (see 6.3):

https://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.32.8520

EDIT: remove code tag from the 20 results to bold the maximum DEAD

[Jean_Luc]

Thanks for the reading and the tests
The dead kangaroos are well handled, they are detected, deleted and recreated.
Was quite a nightmare with the GPU and coalesced memory access

The only things that might happen is that a kangaroo enter in a cycle without having a distinguished point in the cycle but it is rather unlikely.
Handling this in the GPU (with a Floyd cycle detection) will just kill GPU performance.

[COBRAS]

Thanks for the reading and the tests
The dead kangaroos are well handled, they are detected, deleted and recreated.
Was quite a nightmare with the GPU and coalesced memory access

The only things that might happen is that a kangaroo enter in a cycle without having a distinguished point in the cycle but it is rather unlikely.
Handling this in the GPU (with a Floyd cycle detection) will just kill GPU performance.

Jean_Luc,  can you edit code for geting good resualt ? Please.... I very hope Kangaroo will can brute bitcoin...


??

[Jean_Luc]

Jean_Luc,  can you edit code for geting good resualt ? Please.... I very hope Kangaroo will can brute bitcoin... ??

I'm sorry, I do my maximum, but this algorithm is known from long time and it cannot break Secpk1.
May be a flash of genius will hit a day me but for the moment nothing better than O(sqrt(n))...

[COBRAS]

Jean_Luc,  can you edit code for geting good resualt ? Please.... I very hope Kangaroo will can brute bitcoin... ??

I'm sorry, I do my maximum, but this algorithm is known from long time and it cannot break Secpk1.
May be a flash of genius will hit a day me but for the moment nothing better than O(sqrt(n))...

Maybe ETH is more simple and resualtativ will be

Br !!!!! You are great coder and very good man Bro. I will hope you and Etar will psenesnt your enoter result and codes. Very inyresting !!!!!

Etar,  yours method is false resalt generated too, like Kangaroo? 😢

Btc not hacked. This is very bad news for me ((()

[Jean_Luc]

They should be spread with a -(k2-k1)/2 translation, the keys you solved are close to the upper bound, it explains the factor of 2.

I added the translation, looks like better.
I published a new release https://github.com/JeanLucPons/Kangaroo/releases
Thanks to test it

I redid the test above:

Code:

[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^33.12][Dead 1][01:32][352.0MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^32.87][Dead 0][01:19][297.9MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.24][Dead 1][52s][195.1MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^32.62][Dead 0][01:07][251.1MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^33.00][Dead 0][01:27][325.7MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^32.92][Dead 1][01:22][307.3MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.24][Dead 1][01:42][382.4MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.09][Dead 1][01:32][345.0MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.82][Dead 0][01:17][288.6MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.78][Dead 0][01:14][279.1MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.31][Dead 2][01:47][401.0MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^34.06][Dead 5][02:59][672.8MB]
[105.71 MKey/s][GPU 105.71 MKey/s][Count 2^29.98][Dead 0][12s][44.4MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.59][Dead 0][02:09][485.6MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.92][Dead 0][01:22][307.3MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.96][Dead 1][01:24][316.6MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.41][Dead 2][01:54][428.9MB]

[COBRAS]

They should be spread with a -(k2-k1)/2 translation, the keys you solved are close to the upper bound, it explains the factor of 2.

I added the translation, looks like better.
I published a new release https://github.com/JeanLucPons/Kangaroo/releases
Thanks to test it

I redid the test above:

Code:

[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^33.12][Dead 1][01:32][352.0MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^32.87][Dead 0][01:19][297.9MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.24][Dead 1][52s][195.1MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^32.62][Dead 0][01:07][251.1MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^33.00][Dead 0][01:27][325.7MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^32.92][Dead 1][01:22][307.3MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.24][Dead 1][01:42][382.4MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.09][Dead 1][01:32][345.0MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.82][Dead 0][01:17][288.6MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.78][Dead 0][01:14][279.1MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.31][Dead 2][01:47][401.0MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^34.06][Dead 5][02:59][672.8MB]
[105.71 MKey/s][GPU 105.71 MKey/s][Count 2^29.98][Dead 0][12s][44.4MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.59][Dead 0][02:09][485.6MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.92][Dead 0][01:22][307.3MB]
[115.21 MKey/s][GPU 115.21 MKey/s][Count 2^32.96][Dead 1][01:24][316.6MB]
[115.22 MKey/s][GPU 115.22 MKey/s][Count 2^33.41][Dead 2][01:54][428.9MB]

😊😊😊😊Thaaaaank Yoooouu !!!! I wil buy my additional CUDA for testing codes. My first CUDA away from me now.... and aftef I will brute brute brute" brute the wolrld again". Biggest thanks Jean_Luc !!!!

[MrFreeDragon]

They should be spread with a -(k2-k1)/2 translation, the keys you solved are close to the upper bound, it explains the factor of 2.

I added the translation, looks like better.
I published a new release https://github.com/JeanLucPons/Kangaroo/releases
Thanks to test it

Nice job! Seems we were right that the better spread of kangaroos should be applied.
I made the quick test for 16 64bit range keys, and the total time now 10:23 min compared with 16 minutes earlier. Now it is 38-39sec per one 64bit range key!

Code:

$ ./kangaroo -t 0 -gpu in16.txt
Kangaroo v1.3
Start:49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E0000000000000000
Stop :49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5EFFFFFFFFFFFFFFFF
Keys :16
Number of CPU thread: 0
Range width: 2^64
Number of random walk: 2^20.81 (Max DP=9)
DP size: 9 [0xff80000000000000]
GPU: GPU #0 GeForce GTX 1080 Ti (28x128 cores) Grid(56x256) (177.0 MB used)
SolveKeyGPU Thread GPU#0: creating kangaroos...
SolveKeyGPU Thread GPU#0: 2^20.81 kangaroos in 9226.1ms
[512.05 MKey/s][GPU 512.05 MKey/s][Count 2^32.24][Dead 0][12s][759.1MB]  
Key# 0 Pub:  0x0259A3BFDAD718C9D3FAC7C187F1139F0815AC5D923910D516E186AFDA28B221DC 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5EBB3EF3883C1866D4 
[463.81 MKey/s][GPU 463.81 MKey/s][Count 2^33.12][Dead 0][32s][1398.1MB]  
Key# 1 Pub:  0x02A50FBBB20757CC0E9C41C49DD9DF261646EE7936272F3F68C740C9DA50D42BCD 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5EB5ABC43BEBAD3207 
[451.45 MKey/s][GPU 451.45 MKey/s][Count 2^33.22][Dead 1][34s][1492.7MB]  
Key# 2 Pub:  0x0304A49211C0FE07C9F7C94695996F8826E09545375A3CF9677F2D780A3EB70DE3 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E5698AAAB6CAC52B3 
[382.47 MKey/s][GPU 382.47 MKey/s][Count 2^33.70][Dead 1][46s][2086.9MB]  
Key# 3 Pub:  0x030B39E3F26AF294502A5BE708BB87AEDD9F895868011E60C1D2ABFCA202CD7A4D 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E59C839258C2AD7A0 
[386.59 MKey/s][GPU 386.59 MKey/s][Count 2^32.37][Dead 0][22s][827.6MB]  
Key# 4 Pub:  0x02837A31977A73A630C436E680915934A58B8C76EB9B57A42C3C717689BE8C0493 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E765FB411E63B92B9 
[163.64 MKey/s][GPU 163.64 MKey/s][Count 2^30.96][Dead 0][14s][315.7MB]  
Key# 5 Pub:  0x020ECDB6359D41D2FD37628C718DDA9BE30E65801A88A00C3C5BDF36E7EE6ADBBA 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E7D0E6081C7E0E865 
[357.29 MKey/s][GPU 357.29 MKey/s][Count 2^32.26][Dead 0][22s][766.9MB]  
Key# 6 Pub:  0x0322DD52FCFA3A4384F0AFF199D019E481D335923D8C00BADAD42FFFC80AF8FCF0 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5EC737344CA673CE28 
[466.18 MKey/s][GPU 466.18 MKey/s][Count 2^33.14][Dead 3][32s][1415.1MB]  
Key# 7 Pub:  0x02DB4F1B249406B8BD662F78CBA46F5E90E20FE27FC69D0FBAA2F06E6E50E53669 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E38160DA9EBEAECD7 
[455.70 MKey/s][GPU 455.70 MKey/s][Count 2^33.25][Dead 3][34s][1519.5MB]  
Key# 8 Pub:  0x023BD0330D7381917F8860F1949ACBCCFDC7863422EEE2B6DB7EDD551850196687 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E79D808CAB1DECF8D 
[485.68 MKey/s][GPU 485.68 MKey/s][Count 2^32.90][Dead 0][28s][1199.2MB]  
Key# 9 Pub:  0x02332A02CA42C481EAADB7ADB97DF89033B23EA291FDA809BEA3CE5C3B73B20C49 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E54CAD3CFBC2A9C2B 
[485.79 MKey/s][GPU 485.79 MKey/s][Count 2^32.90][Dead 0][28s][1198.0MB]  
Key#10 Pub:  0x02513981849DE1A1327DEF34B51F5011C5070603CA22E6D868263CB7C908525F0C 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E0D5ECCC38D0230E6 
[350.32 MKey/s][GPU 350.32 MKey/s][Count 2^33.99][Dead 3][56s][2542.9MB]  
Key#11 Pub:  0x03D4E6FA664BD75A508C0FF0ED6F2C52DA2ADD7C3F954D9C346D24318DBD2ECFC6 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5EE3579364DE939B0C 
[159.06 MKey/s][GPU 159.06 MKey/s][Count 2^30.92][Dead 0][14s][307.2MB]  
Key#12 Pub:  0x0356B468963752924DBF56112633DC57F07C512E3671A16CD7375C58469164599D 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E7C43B8E079AE7278 
[465.93 MKey/s][GPU 465.93 MKey/s][Count 2^33.07][Dead 0][32s][1343.7MB]  
Key#13 Pub:  0x03D5BE7C653773CEE06A238020E953CFCD0F22BE2D045C6E5B4388A3F11B4586CB 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E8D63EF128EF66B42 
[465.59 MKey/s][GPU 465.59 MKey/s][Count 2^32.99][Dead 0][30s][1274.1MB]  
Key#14 Pub:  0x02B1985389D8AB680DEDD67BBA7CA781D1A9E6E5974AAD2E70518125BAD5783EB5 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E2452DD26BC983CD5 
[429.07 MKey/s][GPU 429.07 MKey/s][Count 2^33.43][Dead 0][38s][1724.7MB]  
Key#15 Pub:  0x0355B95BEF84A6045A505D015EF15E136E0A31CC2AA00FA4BCA62E5DF215EE981B 
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5E7AD38337C7F173C7 

Done: Total time 10:23

Kangaroo v1.3 has better performance for sure! And good thing for Green Peace as Kangaroo v1.3 goes to the target with less "dead animals" 

PS. Later I will test these 16 keys for the wider 80bit range and let you know. Expect it will solve them for 1.5-2 hours instead of 4 hours as it was before.

[Etar]

Confirm that v1.3 has better perfomance! @Jean_Luc Thank you for job!
First key in range 2^80 found in [01:15:13] with 8 dead kangaroo.

Code:

Kangaroo v1.3
Start:49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB4800000000000000000000
Stop :49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48FFFFFFFFFFFFFFFFFFFF
Keys :16
Number of CPU thread: 0
Range width: 2^80
Number of random walk: 2^22.09 (Max DP=15)
DP size: 15 [0xFFFE000000000000]
GPU: GPU #0 GeForce RTX 2080 Ti (68x64 cores) Grid(136x256) (417.0 MB used)
SolveKeyGPU Thread GPU#0: creating kangaroos...
SolveKeyGPU Thread GPU#0: 2^22.09 kangaroos in 25051.2ms
[858.84 MKey/s][GPU 858.84 MKey/s][Count 2^42.15][Dead 8][01:15:13][11359.4MB]
Key# 0 Pub:  0x0259A3BFDAD718C9D3FAC7C187F1139F0815AC5D923910D516E186AFDA28B221DC
       Priv: 0x49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48CB5EBB3EF3883C1866D4
[1331.49 MKey/s][GPU 1331.49 MKey/s][Count 2^38.14][Dead 0][04:43][709.3MB]

[Jean_Luc]

Confirm that v1.3 has better perfomance! @Jean_Luc Thank you for job!
First key in range 2^80 found in [01:15:13] with 8 dead kangaroo.

Many thanks for testing, you should to decrease a bit your grid size, you have a too small dp for a 2^80 range and a large RAM usage.
Could you try with -g 136,128 or -g 68,256 or even -g 68,128, thanks.
This GPU should be quite good at 2^88 range with nominal settings

[COBRAS]

 

Start:49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB4800000000000000000000
Stop :49DCCFD96DC5DF56487436F5A1B18C4F5D34F65DDB48FFFFFFFFFFFFFFFFFFFF
  

Good night Supermans !!! My brain cant sleep becouse this megathread. :-)

Please explain me, why always used this start-stop ranges ? This ranges, as I right ubdersta d NOT FOR ANY publick keys and if I will use publick key not from this thread I will be needed enother ranges Yes or ....

p.s. BEEEEEEEEEG THENKS FOR ALL OF YOU !!!

[MrFreeDragon]

-snip-
Please explain me, why always used this start-stop ranges ?
-snip-

We just use the same public keys and ranges in order to test different devices and help JeanLuc to develop his program. If we use different ranges and keys it will be a little bit difficult to undertand the real improvement of the program performance.

The 16 public keys were generated by odolvlobo just for test purposes earlier in this topic:
https://bitcointalk.org/index.php?topic=5238719.msg54184513#msg54184513

The initial range was only 2^64 width, and now we see that this range is too small for the program (the key is found just for 1 minute or less). So for test purposes Etar decided to make the key range wider up to 2^80 (just change the start and end range).

And at the moment we use that enlarged ranges, together with the same 16 public keys.

If you read that test post by odolvlobo you could find that he made a challenge: "I'll give you a whole day to find the private keys for these 16 public keys. ...  If you can do that, I'll be impressed. I don't think that you can find 4". But, JeanLuc accepted this challenge and developed his own tool on CUDA, and after some days we see that all 16 keys in the range 2^64 could be found just for 10 minutes. There is no need 24 hours for it!

This was the explanation why we use that ranges. However, answering on your question, of course you can use any ranges you want and public keys you want.