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

CUDA/OpenCL testing some algos not working AMD/nVidia #148

Open
Xn0xic opened this issue Jun 5, 2024 · 4 comments
Open

CUDA/OpenCL testing some algos not working AMD/nVidia #148

Xn0xic opened this issue Jun 5, 2024 · 4 comments

Comments

@Xn0xic
Copy link

Xn0xic commented Jun 5, 2024

nVidia 3000 series- cn/gpu, panthera, ghostrider, flex, argon2/chukwav2: Algo skipped due to disabled backend

AMD 5000 and 6000 series- panthera, flex, argon2/chukwav2: Algo skipped due to disabled backend

NVD
AMD

OS: Win 11 Pro

  • For GPU related issues: 3080ti/3070ti 555.85/CUDA 11.4 and 12.5. RX5500XT/5700XT/RX6700XT/6750XT/6800XT Adrenalin 24.3.1

Additional context
Thanks for all the work you have been putting in gents. :)

@Spudz76
Copy link

Spudz76 commented Jun 5, 2024

  • cn/gpu should work in all cases, perhaps you are using xmrig-cuda vanilla from upstream? (must use xmrig-cuda MO-fork)
  • panthera/ghostrider/flex/argon2 simply don't have GPU support at all, normal and expected

might be worth writing ghostrider/flex/argon2 support but that is unclear

panthera is RandomX and none of those, even the supported ones, are worth it

@Xn0xic
Copy link
Author

Xn0xic commented Jun 5, 2024

Confirming I'm using the xmrig-cuda MO-fork from here: https://github.com/MoneroOcean/xmrig-cuda/releases/download/v6.21.1-mo1/xmrig-cuda-v6.21.1-mo1-cuda11_4-win64.zip

Looking back at previous config & log files panthera/argon2/ghostrider all ran for a benchmark during initial setup, but never ran (or even attempted) after the initial run.

"worth it" is subjective, but I do concede that in terms of H/W it is horribly inefficient and usually not a profitable venture. There was a hot minute there in Oct-Dec 2023 where it was actually the most profitable to run rx/0 for ZEPH on GPUs. Right now, kawpow is where it's at, but my GPUs do flip to autolykos2 and CN algos frequently enough that I also using mm.exe to switch between XMRig for rx/ and cn/ and GPU specific miners for kawpow and autolykos2.

@Spudz76
Copy link

Spudz76 commented Jun 6, 2024

Yeah same, but I run a GTX1060 so RandomX is absolutely never worth it H/W wise. Kawpow is barely even actually worth it since I get ~5KH/s pay rate for 140W which is even worse than my crappy old Sandybridge CPUs.

@Xn0xic
Copy link
Author

Xn0xic commented Jun 6, 2024

Ouch. 6700XT with kawpow caps at about 140W for 13ish kh/s equivalent to rx/0 right now. 5500XT is 58W for 6kh/s equivalent. Those are my two most efficient cards. nVidia 3000 series cards are 25% less efficient.

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

2 participants