Calling Team 10
- Thread starter Dantoshou
- Start date
-
- Tags
- coronavirus
Soldato
Yeah I keep having to restart the client to get CPU units, but on the whole it's been ok today.
I'll try restarting the client see if it sends down a WU for me thanks
Considering one company brought online over 6,000 GPU's the issue of supply isn't a you thing. Their servers are just overran by new clients pulling work. What many are doing are running both BOINC and FAH so that their system stay utilized as Rosetta@home is also working on it. GPUGrid.net is talking about also jumping in on the action by having their own application too. So, if you aren't getting enough work, please consider at least Rosetta@home for now.
Soldato
Installed it again after about 10 years. Will try and get it crunching when gpu arrives as I have an HD5870 at mo. No Corvid-19 in the drop down list
Soldato
Ok thanks, that's what it's set to. HD5870 running VERY hot lol - showing it's age.
Soldato
My water-cooled gtx1080 has been sat at 42c for a week now 
Associate
my cpu chews through a WU in about 30mins, when there are any. Takes 270 watts to do it though lol!
Soldato
Ouch lol!
Soldato
Need to manually drill into the GPU config on this one. I have a vega and sure it was trying to use cuda hence error and not processing. Any advanced folders using vega that can tear into this?
most of the time its this:
When a WU is available:
I think I need to set opencl and force it away from auto stuff as it cant config correctly.
Right think I sorted it by setting to defaults. It must not have liked an undervolt.
I still had to get away from the cuda setting it was defaulting to use. Now hopefully I can fold on the gpu!! \o/
Code:
WARNING:WU01:FS01:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
Code:
WARNING:WU00:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
12:38:37:WU00:FS01:Sending unit results: id:00 state:SEND error:FAULTY project:11749 run:0 clone:998 gen:7 core:0x22 unit:0x000000128ca304e75e6a80003bb1e921I think I need to set opencl and force it away from auto stuff as it cant config correctly.
Right think I sorted it by setting to defaults. It must not have liked an undervolt.
I still had to get away from the cuda setting it was defaulting to use. Now hopefully I can fold on the gpu!! \o/
Last edited:
How do you have your cores set up in the app?
I've dropped the clock and voltage on mine, sits at 90-100W at 3.7 GHz/1V.
Soldato
Titan black was sitting at 85c when folding which I didn't like. Added the NZXT baracket to allow my old Kraken CPU cooler to it and it now sits at 52c after continuous GPU folding and silently too!
Far happier now.
Far happier now.
Last edited:
Soldato
Well that didn't take long to escalate, went looking for a GPU and bought another rig, Ryzen 5 3600 with a RTX 2060 super.
Old i7 rig on cpu only was 30000 ppd, new rig is showing 1.5 million ppd
Old i7 rig on cpu only was 30000 ppd, new rig is showing 1.5 million ppd
Soldato
Very nice! At least you're already above me so I don't need to worry about you passing me with my measly 350k ppd!
Can you elaborate? I'm using 100% of my 3960X on BOINC/Rosetta but my GPU (Vega 56) has been idle the whole 24h FaH has been running. It's saying it's available in the advanced config and just no WU, but I'm wondering if it's not set up correctly. I have the latest AMD WQL driver and the default FaH install (no CPU slot, just GPU).
Soldato
Yeah it has a config file, so if you browse to where that text file is I deleted the cuda line. It may be a regular warning but it was a parameter each time being spun that wasnt necessary as its never going to run on cuda. I do get the odd work unit but most of the issues are the servers arent dishing out any WU's because the farms and companies being generous eat up all of them.