Quote from: Firebox on April 05, 2020, 02:11:11 PM

Quote from: alek76 on April 05, 2020, 01:52:32 PM

Quote from: Firebox on April 05, 2020, 01:10:26 PM

Quote from: alek76 on April 05, 2020, 11:39:35 AM

Quote from: Firebox on April 05, 2020, 11:23:27 AM

Unfortunatelly I'm not really expert in this. I'm using 'Visual Studio 2019'.
I have finally succeded to comply vs-kangaroo-hybrid.exe by choosing 'Release' in configuration (before was 'Debug' by default), but without any aadditional settings. Just as it was.
Where do I put this settings "make gpu=1 ccap=20 all"?
Another question here. Runnigng test and found that script goes over 100% i.e. for #50:

Compute capability 2.0 and 3.0 Not supported.
This is normal, the program does not make accurate calculations. Set -p 0 or for bits 50 use cpu version.

Ok. Agree, that could be, but we are triying to test gpu usage, not a cpu )). And to test it quickly I choose lower number addresses in order to see if I will get a propper result in the end.
So far looks like somethis I did wrong:

Result received in ~ 6.5 minutes

Result received in ~ 4.5 minutes

What am I doing wrong? This are the time to spent by CPU.
Could you share you results for #50?

Just when power is set 12 for bit 50, it turns out Hop_module 41 is 1 long jump 2 ^ 41 - it's too far and the range of selected points goes far from bit 50. Accordingly, the search time will be much longer. Also, when a large power - DP parity / DPmodule becomes less than 18, which leads to a large amount of data received from the video card and errors occur, wild and tame files grow quickly. Accordingly, it also affects negatively. The optimum size of the DPmodule is 22 - 24 of these values should be selected power. And if DPmodule is large, then fewer selected points will be saved in files and search time will increase. Therefore, it is installed no more than 24.
+ the comparator starts the comparison once every half a minute. Therefore, such a total time.

All noted. Thanks. Will continue testing.
Those two results that I gave a screenshots were both with p=0. But the one with lesser time was tested with rekey=5000.

To be honest, I didn't test the rekey option at all, but left it just in case. According to the documentation - in a parallelized version of the Kangaro algorithm, all selected points transmitted by clients are stored on the server in one database, where they are then checked for collisions. Therefore, rekey should work. And for point aggregation, you do not need to delete wild and tame files, they are overwritten every time you start, I would recommend commenting out lines ReWriteFiles(); in Vanity.cpp

Code:

if (TWRevers) {
bool F2Upload = TWUpload();
if (F2Upload) {
printf("\n[i] Upload tame.txt and wild.txt ");
}
  }
  else {
//ReWriteFiles();
  }

And put in the directory with the program empty files tame.txt and wild.txt In this case, be sure to install a fixed DPmodule! Set int pow2dp = 22; // Fixed in GPUEngine.cu - This is for bit 110.