Stuck on "paused: waiting for idle" for gpu

[WaKen]

so i ran folding at home couple of days ago and first night i ran it on medium as that is the default and everything ran fine, however i noticed it cranked all my cores up to 100% and gpu ran too although i didnt remember to check % utilization on that which resulted in my water loop getting pretty warm and i got a bluescreen during the night. i know it started using my gpu as it got to 21% on the progression bar thing.
So i decided to turn the speed down to lightand that seems to run my cpu at around 60-68% utlization but it still cranked 2 of my 16 threads (9900k) up to 100% which worries me but it has been running stable. But onto the issues, the web control keeps refreshing a lot, cant use web control which is annoying and for some reason the gpu just wont start working at all. GPU is always stuck on paused: waiting for idle even though i have it set to fold while im working. I even tried to reinstall the program due to these 2 issues but that did not resolve it, ive also tried a solution another member suggested in another thread about being stuck on idle which was to turn it to only work on idle and back to while working. that did not make my gpu start working. Im currently running an RTX 2080ti with nvidia driver 442.59.
I would love it if the settings of light-medium-full werent cranked so high, medium is using my 9900k at 100% cranking everything up to 5ghz which it turns out it cant run for extended periods of time. i dont feel like adjusting my overclocks just because i cant run whats effectively a long stress test for my system.
it's frustrating that only my cpu is doing any work since a gpu should bring a lot more horse power to the table, i was hoping it was folding on COVID-19 as that's what im currently hoping to help with. But on the cpu side its not one of the options in the drop down menu, it says im working on project 14303 without giving me any indication of what that is. Had to look it up to figure out it is cancer, i dont mind helping on cancer but i would love it if i could set my gpu to work on Covid-19.
Any help to get my gpu up and running is appreciated.

[muziqaz]

I'm not certain if I'm correct, but GPUs are very hard to keep idle, if you are browsing internet, browser is accelerated by GPU. Maybe fahcontrol assumes GPU is doing work even when you think its idle. As I said I'm just thinking out loud.
Now if your system is crashing while under medium work, consider dialing down your OC, as folding@home is one of the best ways to see if your system is 100% stable. Many of the gamers consider their systems stable if they don't crash while playing for few hours. Folding is completely different beast and roughly equals or even tops Prime95.
CPU work units use AVX256 instructions, which on Intel systems forces CPU to heat up, and Intel implemented some AVX load offsets to counter that, basically they have different clocks for SSE loads, and AVX loads. In BIOS you might have seen such an option.
GPU work units use EVERYTHING GPU has to offer, while todays games use like 50% of GPU capabilities with broken or buggy graphic engines.
Now if you want to contribute towards COVID, "Any" option in Fahcontrol>Configure>Advanced section will prioritise COVID work units. Keep in mind, researchers are working around the clock to add more work related to COVID, but you are so many, and they are so few

[kiore]

If your CPU getting too hot you can adjust the number of threads manually using the advanced control setting. For example I have a Ryzen 8 core 16 thread CPU but set it to 12 threads only, it still does serious work but is not so hot. Your GPU maybe be waiting for work assigned as so many new donors have joined the assignments servers have been overloaded with requests.
COVID-19 units are on both GPU and CPU.

[bruce]

The choices in WebControl are very limited. If you want precise control, use FAHControl and adjust your settings carefully.

GPUs do not have speed choices unless you make changes in the drivers. As far as the slider in WebConrol, the GPU is either folding at 100% or it's not. The OS treats it like a coprocessor,

Using FAHControl, you can adjust what load FAH puts on your CPUs. Your choices will be based on the number of threads (CPUs) your system has, from 0, 1, 2,.. up to all but one of them. One CPU thread is reserved to move data to/from each GPU.

All WUs use either N of your CPU threads or they use 100% of the GPU plus one CPU and you can elect to use only CPU WUs or only GPU WUs. Initially all COVAID projects were processed on the GPU. Now you choose to run other COVAID projects on the CPU.

If you are unwilling to reduce your overclock, you've got some choices to make, because FAH is designed to use ALL of the resources you're willing to configure for it to use and it tends to be more power hungry than many of the standard overclocking benchmarks.