Folding Forum

@pannu: Would you be willing to be pulled into the core dev testing slack? If so, DM me your email address and we'll invite you to help track this down.

Thanks so much for bearing with us!

~ John Chodera // MSKCC

> It would be nice if FAH would show "Please update nvidia drivers to >= v4xx.yy" instead of CUDA_ERROR_INVALID_PTX only.

We weren't sure this was the cause of the error until recently, but we can add this message in the next core release!

Thanks!

~ John Chodera // MSKCC

> When did the cuda update in folding drop again?? T

core22 0.0.13 with CUDA support rolled out on Mon 28 Sep for most folks (though BETA users had it more than a week earlier).

Is just this project giving you trouble, or all of them?

~ John Chodera // MSKCC

Thanks for the update, and let us know if you end up with something like this again!

~ John Chodera // MSKCC

> Core 22 v. 0.0.13 is apparently released to non-beta users too.

Whoops! This made it out of the testing lab accidentally. We're still doing a lot of testing, so nothing official yet...

~ John Chodera // MSKCC

Just wanted to chime in here: We hope to be able to proceed with this refinement of GPUSpecies soon, but don't have a concrete timetable. Getting good benchmark data out for the comprehensive 1710x benchmark projects is our first priority! After that, it should be straightforward to begin to automat...

We have some bugfixes for AMD Cards coming in the next core release that may help here!

~ John Chodera // MSKCC

> Than why FAH doesn't have option to send data in bursts? As even amateurs could make simple utility to prevent gpu from heating down and heating up! As that was the problem! We are not talking here about anything overly complicated. And problem would be solved... We've explored this in the past, a...

@silverpulser: If you want to experiment with this, it would be better to do it in the #core22 development slack I invited you to since we have an NVIDIA rep working with the driver team who can help guide you through specific tests!

~ John Chodera // MSKCC

Glad the driver rollback worked!

We've been working with some good folks at NVIDIA to better understand these failures, so if you have complete details on how to reproduce the issue, that would be super helpful!

~ John Chodera // MSKCC

I'm glad you posted! We've been seeing a few cases of this error: clCreateContext (222) This isn't a valid OpenCL error code, and likely means there is something going wrong with the NVIDIA driver. Which driver version do you have installed? Is there any chance this was updated recently? Does reboot...

I'm showing the stats credit for 13433 (SARS-CoV-2 Mpro monomer) is 70660, and for 13737 (Mpro dimer) is 118000. Are you seeing the base credit is _lower_ for 13437 than 13433?

> The base points is actually lower for 13437 than 13433 so logically it should be a little bit quicker ... with these PRCGs someone will be able to see if anything obviously odd is going on.

Whoa, that may be a mistake. Looking into this.

We're up to ~2/3 of the core22 assignments now, so I think this is helping.

Thanks again for the heads up!

~ John Chodera // MSKCC

> In the past few days I’ve been getting some Moonshot WUs that take over 4 hours to complete on either of my GPUs (one RTX 2060, one RTX 2060 KO, no overclocking) — an hour or more longer than what was usual earlier. I’ll take that as a sign that WUs can and will be adjusted for the new GPUs due to...