Mekhed
This is the only Bad WU I've had in the last 2 days. Not a core 21 ...No bad states
 
22:54:05:WU01:FS01:0x18:Completed 1750000 out of 5000000 steps (35%)
22:54:15:WU01:FS01:0x18:ERROR:exception: Error downloading array energyBuffer: clEnqueueReadBuffer (-5)
22:54:15:WU01:FS01:0x18:Saving result file logfile_01.txt
22:54:15:WU01:FS01:0x18:Saving result file log.txt
22:54:16:WU01:FS01:0x18:Folding@home Core Shutdown: BAD_WORK_UNIT
22:54:16:WARNING:WU01:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
22:54:16:WU01:FS01:Sending unit results: id:01 state:SEND error:FAULTY project:10477 run:1 clone:9 gen:171 core:0x18 unit:0x000000e3538b3dba548b25d790a32fc0
22:54:16:WU01:FS01:Uploading 2.82KiB to 140.163.4.234
22:54:16:WU01:FS01:Connecting to 140.163.4.234:8080
22:54:17:WU01:FS01:Upload complete
22:54:17:WU01:FS01:Server responded WORK_ACK (400)
22:54:17:WU01:FS01:Cleaning up

scott@bjorn3d
I am thinking Stanford screwed up GPU compute on Maxwell products.  My original Titan can fold all day long core 21 with no issues.  I have core 21's go crazy every now and then with GPU usage and power jumping up and down all the time.  I am using MSI's Afterburner to watch my GPU's as it gives better live logging compared to EVGA's.  I never have a problem on a core 15,17 or 18.  Only core 21 gives me issues on the Maxwells.  When all works fine I am getting 1 million PPD with a Titan in one machine and a GTX 980TI SC and 980 SC in my main machine.

kougar
Slightly off topic but related...  after I updated Windows 10 to the new 1511 November build it somehow broke the OpenCL implementation that FAH uses. I'd get OpenCL.dll missing errors trying to run Core 21 projects on a Titan Black. Reinstalling the drivers fixed it, but it was strange that I had to do so at all...

Vijay Pande
We’ve discovered that the new FahCore_21 is producing more errors than we consider acceptable for some clients.  The error rate seems to depend on several factors but most noticeable is that it doesn’t work well with second generation Maxwell GPUs.  A few projects have made their way through Advanced testing have been distributed to everyone under the default “FAH” client-type setting.  To allow donors to limit this exposure, those projects have been reclassified as “Advanced” which is appropriate for a FahCore that is still under development.
As has always been the case, the “Advanced” setting will give you access to newer projects which may have a higher error rate.  It is our intention to provide only the safest assignments with the default setting or you can choose to configure your system to run these advanced projects depending on how frequently you encounter these errors.
These conditions are expected to improve as new projects, new versions of that FahCore, or new versions of the drivers incorporate whatever fixes are required.  In the meantime, Work Units which are completed successfully allow scientific research to progress toward even more challenging projects than we’ve done so far.
 

troy8d
https://folding.stanford.edu/home/issues-with-fahcore21/
 

Vijay Pande
We’ve discovered that the new FahCore_21 is producing more errors than we consider acceptable for some clients.  The error rate seems to depend on several factors but most noticeable is that it doesn’t work well with second generation Maxwell GPUs.  A few projects have made their way through Advanced testing have been distributed to everyone under the default “FAH” client-type setting.  To allow donors to limit this exposure, those projects have been reclassified as “Advanced” which is appropriate for a FahCore that is still under development.
As has always been the case, the “Advanced” setting will give you access to newer projects which may have a higher error rate.  It is our intention to provide only the safest assignments with the default setting or you can choose to configure your system to run these advanced projects depending on how frequently you encounter these errors.
These conditions are expected to improve as new projects, new versions of that FahCore, or new versions of the drivers incorporate whatever fixes are required.  In the meantime, Work Units which are completed successfully allow scientific research to progress toward even more challenging projects than we’ve done so far.
 

 
It seems like your efforts have paid off and our voice has been heard.

Scarlet-Tech

troy8d
https://folding.stanford.edu/home/issues-with-fahcore21/
 
 
It seems like your efforts have paid off and our voice has been heard.

This is INCREDIBLE. I made to share it on folding forums so it is seen on our thread also.

Drazhar
Speaking of issues with Core21 and GTX 980s, anyone else notice a total desktop slowdown after a certain point while folding Core 21 projects? It immediately goes away if I pause folding and takes a while to come back on resuming, but it's not showing any errors in the log.

Mekhed

Drazhar
Speaking of issues with Core21 and GTX 980s, anyone else notice a total desktop slowdown after a certain point while folding Core 21 projects? It immediately goes away if I pause folding and takes a while to come back on resuming, but it's not showing any errors in the log.

Yes I've had that with my 980Ti. ...are you using the overvoltage tool in PX16?