Yeah it was weird but I was running out of space...was gobbling up RAM and I'm assuming pagefile as well but when I moved the app to a separate drive it is now running at 500Mkeys/s

As for the "every 5th" I now understand that this algo requires the public key and the other BTC puzzle programs just brute force all the private key options in a range

Ok, still trying to wrap my head around the concept But getting there. Hats to those of you here who understand all this

So I got it to run.
1080ti on Windows
Bat file >> Kangaroo.exe -t 0 -gpu -gpuId 0 -ws -w save.work -wi 30 input.txt

Starts off at 200Mkeys/s and then continuously drops

Also why does this algo seem to only work for bit ranges that are multiples of 5?

 
Thanks. So the idea is that the tame herd is a series of numbers that are  increasing in a constant way and the wild herd is variables that are "randomly" moving forward...when a collision occurs, how does that determine the public key?

Also, I did get the prog to run but this is what I get:

This output makes no sense to me as its unclear as to how well my GPU is performing. What is "dead" and why am I only using 12MB of expected RAM when I have 11GB?

Running windows, but have a pc I can install Ubuntu on (have the USB ready

Sorry I cant add to the discourse but would really appreciate if you guys could help me out
1) Is the kangaroo algo essentially looking for the x,y coordinates on the curve that make up BTC key? All I could gather was that essentially 2 random points are selected and plotted and if an eventual collision occurs you have the key?

2) I tried to run the program with my 1080ti but couldnt make heads or tails of the settings (hence my previous question to try and understand what it all pertains to)...could you help me out with a config that would be adequate for that gpu?

Thanks

Thanks, but I am trying to learn a little bit about how workloads are distributed to GPUs as well as optimizing my gpu...so could you explain how you decided on those values or was it trial and error with a similar gpu?

Hello,
Can anyone explain the relationship between -b -t and -p and GPU work groups? I tried doing some research to try and optimize this but couldnt and I dont have the coding know how to code a simple benchmarking utility that would test the ranges.

I assumed that if I learned a little about workgroups as well as knowing the compute units for my gpu (I have an RX 580 so it has 36 CU with 2304 SMs) I could better determine a good configuration.

Thanks

Hello,
I was hoping someone would be kind enough to simplify what the program is doing. What I gathered was that it using the Kangaroo algo which essentially picking 2 random points at set intervals and checking for a collision. I think I started trying to read the start of this thread but was totally lost. I have 1 1080ti that I like to tinker with...have made vanity addresses, messed with Bitcrack for the 32btc puzzle but those were easy to understand. I dont like running things unless I learn something

Also I tried running it but the gpu never really started processing anything and couldnt make heads or tails of the output on the CMD

Any simplified explanation would be greatly appreciated so I can use that to then start learning more and tweaking , etc.

Thanks