zophar
Well, after a 4 day test I can say a 1080TI is good for 1.65 million PPD on core 22. I've had 4 days with only a 3,000 point spread. Definitely a pick me up with the higher year 12 point requirements and only having a single card. Also with the latest driver I've had no unit failures. Now let's see if they leave the points alone when it leaves beta or if it drops.

Over 400 WU with no BAD or Failed WU

Wow, my 2080TI's are eating these core 22's like Fluffy eats tacos 
 
Hella nice PPD and no failed units yet.  Granted I've only done like 4 on them.

STR1D3R_2
I was late to the party also

Holdolin
Wow, my 2080TI's are eating these core 22's like Fluffy eats tacos 

chomp chomp Like adding a free GPU
 

Holdolin
Hella nice PPD and no failed units yet.  Granted I've only done like 4 on them.

 
Been running for 7 days now and I'm not sure if this was a core 22 issue or not but I had to hard restart one rig yesterday morning. Looking at the points log, it seems to have locked up just before midnight the night before. After a reboot it completed wu's and hasn't had another hiccup since. That rig sits closest to the window that I left open with a fan blasting air in and that night it dropped down into the low 20's maybe even the mid teens rapidly after a warm day and there was a bit of frost on the plexi that supports the clc cooler sooooo.
"Check yer rigs", frequently

I have ssh terminals open to all my folders and watch them from my workstation.  I use the fah monitor to keep an eye on production and the ssh tunel to keep an eye on tems/load.
 

Whelp, my 2070's don't like 'em at all. Cosmo was generating more errors that getting work done so I removed the beta flag.  I waited to try it on Wanda till this morning (she has a single 2070) and would go to 4% before detecting a bad state.  Fearing returning bad results I was debating on just dumping the wu then she did it for me lol.  The 2080tis are just fine though. Still no errors.  Jorgan and his pair of 1080tis kicked about 2 bad state errors over night but seems to be doing ok.  So the long and short: the 2080tis (at least for me) are about the only cards flawleslly folding the core 22s.

blackpanther26
I think they are not distributing this core at this moment. I want to see if this new client will allow my 2080 ti to use 100% of the GPU rather than the 60 to 70% it's currently using with FAHcore 21

My cards are all crunching them as we speak.  Do you have your beta flag set?

It moved back to advanced. I never set anything but pause on start and advanced.

That or there are leftover work units in advanced to go through.
I suppose there could be glitches if advanced units run out to get regular units.

Up to over 400 WUs on my GTX 1080 Ti Card and no Failed or Bad WUs, Still using the Beta Flag. 

ipkha

bcavnaugh
Up to over 400 WUs on my GTX 1080 Ti Card and no Failed or Bad WUs, Still using the Beta Flag. 

Good for you.
My split rig with 1080ti and 1 2080ti reboots or hard freezes every so often with these units. This is without any crunching going on or overclock on the GPU. Trying with 80% power on the 2080ti to see if it helps.

Another reason Not to Mix Graphics Cards being that one Driver for Both no longer works in the best interest IMO.

Yep! and why I had to Drop from 2050MHz to 2037MHz on my GTX Cards.

scott@bjorn3d
I split my rigs due to overloading the backup ups units.  Funny thing is my PPD went up over 600,000 with doing that.  2080TI main card and 1080TI FE as secondary in main rig, 1080TI FTW main card and 2080TI secondary card in living room.

I just split my PSU and UPS Units.
But I do not Mix Different GPU Generations in the same host.

ipkha
I only have 2 rigs up and running. I put my 2 1080ti FTW Hybrids in the older system. I have a 1080ti black hybrid and 2080ti black hybrid in my main rig with 128GB ram. It's my main gaming/folding rig. At least until core 22 started these weird reboots. If split generation card rigs are a problem I'll have to sell a 1080ti and invest in another 2080ti.

What size power supply do you have?  I know I had to split them up as the where taking down my 1500VA UPS with 2 2080TI's in one box folding core 22 with power levels set to max for the cards.  With the boxes spit I can run max power and get better clock speeds on the 2080TI's that way.

1600W T2. 85% power rating lasted 12+ hours. I think it's just clocking too high at base 100% power target and glitching the calculation. As it's the primary display driver that can crash the system. I'll find a good target and leave it be.

Chris21010
are they really that much more sensitive to OC? well, even if they are a highe GPU utilization will offset the lower clock speeds to reach a stable point.

For reference, most of my rigs weren't working well IF at all with Core 22 and NONE OF THEM ARE OVERCLOCKED.
It would appear that little or no testing was done with multi-GPU rigs, or with the 1050 and 1050 ti cards - the ONLY rig I had that didn't crash out and lose GPUs constantly to bad work units and HANGS due to the bad work units was one rig with a 1080 and a pair of 1070 cards.
It's NOT a "OC" specific issue - though their comment about "factory overclocking" is curious, as pretty much EVERY card EVERY manufacturer other than NVidia itself puts out has some level of factory overclock, and if this core was sensitive to THAT it's going to keep having issues.
 

After the problems with not getting GPU WUs assigned (with no or advanced flag) a few days ago I decided to try the beta flag and see how my GTX 980Ti and 1070 (both SC models) would run on the Core 22 WUs.
I've started with both cards setting back to stock (SC) clocks and wanted to test first with the older 373.06 driver, this is still the best performing driver for Core 21 WUs (but I had already downloaded the 391.35 driver, recommended by bcavnaugh).
 
Both cards are running well with this older driver (I did not had a failure yet), but I had to downclock the GTX 1070 to -90 MHz (running at 1847 MHz now), to prevent getting a warning sometimes.  The GTX 980Ti runs without any issue, set it to +50 MHz now, running at 1353 MHz (was running at 1383 MHz on Core 21 WUs).
My total PPD increase is ~25%, the GTX 1070 is ~20% and 980Ti is ~30% (but my 980Ti always underperformed ~10% on Core 21 WUs).