I poke back into this every now and again but just can't seem to make any headway. I've rebuilt my machine with new hardware a few times since it stopped working (over a year ago now) and also a new OS. On the latest nVidia drivers always. Latest BOINC always etc. I really like GPUGrid because it, well at least used to, mean I could set all my CPU cores for WCG and then use my GPU for GPUGrid. I saw in the log it said my online pref's weren't good so I changed them to what looks like a pretty basic but beefy amount of disk/ram/cpu/gpu but still getting the error message below.

10/6/2016 6:47:02 PM | GPUGRID | Your current settings do not allow tasks from this project. To fix this, you can change Project Preferences on the project's web site.
10/6/2016 6:47:02 PM | | Suspending computation - computer is in use
10/6/2016 6:47:02 PM | | Suspending file transfers - computer is in use
10/6/2016 6:47:57 PM | GPUGRID | General prefs: from GPUGRID (last modified 06-Oct-2016 18:42:13)
10/6/2016 6:47:57 PM | GPUGRID | Host location: none
10/6/2016 6:47:57 PM | GPUGRID | General prefs: using your defaults
10/6/2016 6:47:57 PM | | Reading preferences override file
10/6/2016 6:47:57 PM | | Preferences:
10/6/2016 6:47:57 PM | | max memory usage when active: 24518.77MB
10/6/2016 6:47:57 PM | | max memory usage when idle: 24518.77MB
10/6/2016 6:47:57 PM | | max disk usage: 50.00GB
10/6/2016 6:47:57 PM | | don't compute while active
10/6/2016 6:47:57 PM | | don't use GPU while active
10/6/2016 6:47:57 PM | | suspend work if non-BOINC CPU load exceeds 25%
10/6/2016 6:47:57 PM | | max download rate: 7000003 bytes/sec
10/6/2016 6:47:57 PM | | max upload rate: 2000005 bytes/sec
10/6/2016 6:47:57 PM | | (to change preferences, visit a project web site or select Preferences in the Manager)
10/6/2016 6:48:12 PM | GPUGRID | update requested by user
10/6/2016 6:48:12 PM | GPUGRID | Sending scheduler request: Requested by user.
10/6/2016 6:48:12 PM | GPUGRID | Not requesting tasks: don't need (job cache full)
10/6/2016 6:48:14 PM | GPUGRID | Scheduler request completed
10/6/2016 6:48:14 PM | GPUGRID | Your current settings do not allow tasks from this project. To fix this, you can change Project Preferences on the project's web site.

I've tested this with "turn off on activity" both on and off many times. Doesn't seem to matter.

Is it possible it's defaulting to my D:\ which is low on space instead of C which has a fair amount? That wouldn't seem to match the error message though. FWIW I did in fact change the project settings through this site and on the BOINC client.

Edit: On a whim I turned off the suspend options to see what it'd do with my increase in disk space and man, WCG went to town on downloading new work while GPUGrid still said Job cache full when I have no WCG work.

Just basic things, but when you open BOINC and "View" as "Advanced View" and set under "Activity" to "Run Always", "Use GPU Always", and "Network Activity Always", does it start to look for work or have you already done that? Then under "Options", all the "When to suspend" options are unchecked and the "Other" options have "store" options above zero, even if it is just "0.02"? You may not get work, but it may eliminate your not being able to get work. I personally also set the BOINC Manager to be able to use 100% of anything it needs, except 95% of the swap file (because 100% swap file usage can be problematic, even when the OS is managing it to grow as needed.) After all, it is most likely not to use 100% of any of it, but may read less than it needs from settings set lower than it thinks it needs and reject getting work based on a false reading of "expected" resources.

Like I said, just the basic things.

Also, now that I look, there is even more basic that all that. Doh! I am not sure which system you are referring to, but none of the current systems on your list (current as in ones that have tried to contact GPUGRID within the past 30 days, or past year for that matter) seem to have an NVIDIA GPU installed.
https://www.gpugrid.net/hosts_user.php?sort=rpc_time&rev=0&show_all=1&userid=117564
You might want to check on why there is no NVIDIA GPU seen as installed on the system to find out why the system is not allowing itself to receive work for one.
____________
1 Corinthians 9:16 "For though I preach the gospel, I have nothing to glory of: for necessity is laid upon me; yea, woe is unto me, if I preach not the gospel!"
Ephesians 6:18-20, please ;-)
http://tbc-pa.org

I am not sure which system you are referring to, but none of the current systems on your list (current as in ones that have tried to contact GPUGRID within the past 30 days, or past year for that matter) seem to have an NVIDIA GPU installed.
https://www.gpugrid.net/hosts_user.php?sort=rpc_time&rev=0&show_all=1&userid=117564

10/6/2016 6:47:02 PM | GPUGRID | Your current settings do not allow tasks from this project. To fix this, you can change Project Preferences on the project's web site.

You need to edit your GPUGrid web preferences, again, to make sure that "Use NVIDIA GPU" is set to "Yes", "Use Graphics Processing Unit (GPU) if available" is set to "Yes", and possibly also set "Use CPU" to "Yes" and "Use Central Processing Unit (CPU)" to yes.

Also, you need to make sure that applications are selected to be worked on. The default settings, which you could use, are "Run only the selected applications" set to "(all applications)" (all checked), and "If no work for selected applications is available, accept work from other applications?" set to "Yes".

If you're still getting this "Your current settings do not allow tasks from this project." message, then please do us a favor, and post your web preferences. You're not making it easy for us to troubleshoot, currently.

================================================================
10/6/2016 6:48:12 PM | GPUGRID | Not requesting tasks: don't need (job cache full)

You clicked Update to contact GPUGRID, but BOINC decided it didn't need work, because "(job cache full)", meaning that you already have enough work in your local cache to keep you busy.

You should do some research into finding out what "Store at least x days of work" and "Store up to an additional y days of work" actually mean, as they will be relevant to whatever plans you have to store work locally.

And if you want even more detail on work fetch information to be displayed in the Event Viewer, you can turn on the "work_fetch_debug" flag, in the Options -> "Event Log Options" menu.

More details on those logging flags can be found here:
https://boinc.berkeley.edu/wiki/Client_configuration

Good luck!
Jacob

I definitely had it set up that way so uninstalled, reinstalled without running as a service and still am encountering the issue. Going to try some other suggestions now.

Edit: Scratch that, somehow it still installed as a service. Reinstalled again and now it's detecting the GPU but said there's no new work, but no other error messages. Going to just let it run on default settings now and see if it starts working or gives me more errors to work with.

Reinstalled again and now it's detecting the GPU but said there's no new work, but no other error messages. Going to just let it run on default settings now and see if it starts working or gives me more errors to work with.

Reinstalled again and now it's detecting the GPU but said there's no new work, but no other error messages. Going to just let it run on default settings now and see if it starts working or gives me more errors to work with.

Glad you got it working. Work is very spotty right now. The server shows 2 short WUs and no long. If both types are enabled in your project preferences, you'll probably pick one up soon.

Yup! All up and running. Been running BOINC for so long I never realized that the GPU requirements changed. I've always run as a service so that when my computer updates and restarts and I'm not around it'll just start processing again.

Thanks for the help!

Just fyi:

Windows 10 has a feature, now, that can automatically log you back in and lock, after restarting for an update. It's in Settings -> Update & security -> Windows Update -> Advanced Options -> "Use my sign in info to automatically finish setting up my device after an update."

Man, some IT guy I am. Of course I'd never use that feature in my work environments.