Folding Forum

So I've gotten several more of these now, and I'm pretty convinced that the memory fixed it for most. My computer did get a 13405(545, 58, 3) just now though, and it's running at well over 3 min/frame for only 513kppd. Not sure if this is useful feedback, but for some reason this particular run seem...

Yeah, cooling is definitely a huge part of it. Part of the reason I suspect my 1080ti runs so well at 2050MHz is because it's water cooled, and stays under 62-63C even at 300W, and normally when folding, it stays at 47-53C.

...But I've given up on overclocking to avoid wasting any more processing time. As I said on the previous page, FAH does not support overclocking. FAH designes their software to take advantage of every bit of the hardware that is designed into it. Call it a benchmark if you want to, but it's highly...

After experimenting with overclocking over the last couple weeks, there is a clear correlation between clock speed and getting BAD_WORK_UNIT (114 = 0x72) errors. Even an increase of only 130 MHz on GPU causes starts causing errors sometimes. 'Only' 130 MHz - WTF! Days of that sort of overclocking a...

Makes sense. In games though, you're running faster - note that the current mem clock is 5006, but the GPU clock is 5505, so in P0 (which is the performance state in games), you'll be running at 5505 instead (which is a normal clock for a factory 1080ti). It does look like there's a bit of a factory...

1251? That sounds a lot like P2 - even on stock clocks, a 1080ti should be sitting at 1375 or so. Might be worth double checking with Nvidia Inspector? I know on default drivers, it forces P2 on compute workloads.

Also, looking through my logs, I have had a few of these fail as bad work units. I'm pretty sure this isn't my system, since no other WUs have failed in similar ways, and my understanding from the other threads are that this is kind of expected with this project. Here's a log of all the 13404s and 1...

> After playing around with my clocks and settings, it does seem to be pretty memory bandwidth bound. Pushing memory clocks up from 5000 to 5600 remedied most of the PPD difference, though I haven't monitored other WUs to see if I get a similar gain there. Interesting! Thanks for this---that may he...

After playing around with my clocks and settings, it does seem to be pretty memory bandwidth bound. Pushing memory clocks up from 5000 to 5600 remedied most of the PPD difference, though I haven't monitored other WUs to see if I get a similar gain there. Definitely glad to see the researcher's respo...

Interesting. I just grabbed yet another 13404, so I'm trying pausing my CPU folding right now just to experiment and see if maybe I wasn't leaving enough free CPU power to fully feed the GPU, but I'm glad to hear I'm not the only one with odd experiences here.

Also, if it matters at all, here's my current log for 13405 (I just got another one after the 13404 finished): 18:22:30:WU01:FS01:Started FahCore on PID 9440 18:22:30:WU01:FS01:Core PID:19368 18:22:30:WU01:FS01:FahCore 0x22 started 18:22:31:WU01:FS01:0x22:*********************** Log Started 2020-05-...

My GPU has grabbed a couple of these recently (3x 13405 completed successfully, 1x 13404 complete, 1x 13404 in progress), and I'm seeing PPD values all the way down around 950k for 13405 and 920k for 13404. For reference, on other Core 22 projects, I've seen PPDs more in the range of 1.2-1.4M (recen...