Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Quirks, Could Become Advantageous. #151

Open
ForbiddenSoul opened this issue Mar 19, 2015 · 0 comments
Open

Quirks, Could Become Advantageous. #151

ForbiddenSoul opened this issue Mar 19, 2015 · 0 comments

Comments

@ForbiddenSoul
Copy link

First off would like to thank everyone involved with the software. Good job boys and girls good stuff!

Secondly I have noticed some quirks that any miner can use to their (dis)advantage, and the authors can take into consideration to possibly improve the software.

I pool mine, because I'm a hack and have no idea WTF I'm doing, with a GTX550Ti (basically a factory overclocked GTX550).

The first quirk I noticed is in Scrypt mining Litecoin.
With this configuration:
"%~dp0cudaminer.exe" -l F4x16 -o localhost:9332 -u user -p password (basic stuff)
I hit around 58-60 khash/s.
The quirk is when I run another application that want's to use the GPU, in this case, Counter Strike Global Offensive. The game runs flawlessly with a steady 60FPS, and my hash rate increases to 70-76 khash/s.
I believe this is just because the cards clock speed is dynamic and won't run at full power until enough demand is put on it. I supposed you could manually set the clock speed and get the same result. However being able to play video games at a 0% decrease in performance, makes me think that the miner is missing out on TONS of potential, something to think about. (Want to mine leave a game running in the background your has rate will increase by 17-21%)

The second quirk I noticed is in scrypt-jane mining yacoin.
First off I'll say I can not play games or do anything else really while the miner is running, and there is a khash/s loss when opening a game.
Second the most khash/s at the current n factor I was able to get was 0.13 khash/s, I'm not sure if that is good bad or otherwise. It was using this configuration:
-a scrypt-jane:YAC -o localhost -u user -p password -L8 -l f3x18 -b 4096 -C 0 -m 0 -H 1 -i 1
Setting -H 2 (default) I was getting 0.04 khash/s
Setting -H 0 I was getting 0.07 khash/s

(OK, I just doubled checked some things, and this isn't so much a quirk as a performance boost, I think, maybe someone else can shed some light on it)
Using the above configuration I was getting a ton of failed to validate on CPU errors, and was producing very little shares compared to using -H 0 getting 0.07 khash/s with no validation errors and more shares than -H 2.

The sort of quirky part is with separate CPU miners. Running without CUDA miner present, or with -H 2, my CPU miner produces 28 hash/s (I believe 0.028 khash/s please correct me if I'm wrong). With CUDA miner running on -H 0 the CPU miner has rate drops to 10 hash/s (not bad considering -H 0 gives CUDA miner a 0.03 khash/s boost), but the CPU miner is submitting shares more often than it used to, and submitting them at the same rate as CUDA miner (5 on both CUDA miner and the CPU miner in the time it took to write this). So if you want to mine, you can increase share rate of both miners by setting -H 0.
Maybe the -H 1 can be tweaked to not produce validation errors, or possibly it is just my set up, and maybe you could get even more power out of it by combining the CPU miner somehow. (double the share rate)

So those are the quirks I've noticed. I hope something useful comes out of this and thank you all once again!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant