Folding Forum

i was just wondering if F@H is still having trouble getting WUs out to the GPUs.
I read that small molecule screening simulation was just realsesed but i havnt gotten any WU for the last 3 hours for my 2080 Ti
I also realized that my rx 580 gets more WU sent to it and its always running, why is that?

Also my client says that im getting around ~2 million ppd but on the website it only gave me around ~400000 ppd

Yes.

Within the last 4 weeks, Folding at home has 10 times the number of volunteers trying to fold.

In the same 4 weeks they have beefed up the infrastructure so 6 times the work is getting done. (This is faster than I thought they could, a great many companies volunteered servers)

So if F@H had 50% more servers, they might have enough to serve all the new folders. Over all, we are doing much more than a month ago, but individually, we sometimes wait for work.

The server has rules about what Work Units can go to what Graphics cards. Your 2080 ti will do wonderfully on a huge Protein, but will be wasted on a small one. The RX 580 may do better on smaller Proteins, I have to admit I don't keep up on the AMD side as much.

https://www.techpowerup.com/gpu-specs/g ... 0-ti.c3305
https://www.techpowerup.com/gpu-specs/r ... -580.c2938

I notice that the GTX 2080 ti has 2000 more cores than the RX 580, so would be better suited to proteins twice as big.

Have you recently set up a Passkey? The client starts estimating Points with a Passkey immediately, but they start being awarded after 10 Work Units. (Also the client estimates based on being able to get work all day)

Ohh thank you so much for the clarification
I knew about how companies were donating servers to F@H but i didn't know that that was still not enough for the new folders

And yes i set up a pass key yesterday so i guess it still hasn't reached 10 WU requirement yet.

Also i feel like this is a dumb question but i think my computer is only connected to only a few servers out of the 29 for getting WUs. Idk if thats true or not but.
https://apps.foldingathome.org/serverstats

Servers have been donated, but it takes time to setup, configure, and then load projects onto them. Also periodically they have to offload results, the server storage gets filled up fairly quickly.

In addition to what has been stated above, some GPU Projects can be exclusive to AMD GPUs while others are exclusive to Nvidia GPUs and others can run on either. Thus, seeing differences between GPU Vendors is normal.

Another item to mention is that under heavy load, the Stats system stops processing data. Thus, you entire credit may not show up until the Stats system has had a chance to catch-up. Do note that once you have successfully upload your completed WU to the F@H Servers, credits will not be lost in 99.9% of situations.

You system is only talking to those F@H Servers that can provide you with the WU that best matches your system. Thus, it is normal to be connecting to a handful of F@H Servers for most of the time.

thank you guys!

I have one more question for now.

When my GPUs are looking for work units and tries an attempt and fails to get a WU, the timer for the next attempt keeps going up til hits hours. I would have to pause the GPU manually to reset the attempt timer.

So is there a way to change it so that the timer doesn't keep going up as you accumulate attempts or is there a way automate what ive been doing of manually resetting the timer.

I am not a member of the FAH team, just a regular folder but as far as I understand it:

This behaviour is by design to protect the servers and is fixed at this time … pause/fold is a manual intervention that resets the timer but the team would prefer this is only used once the retry timer is hours rather than minutes … I am sure that programmers will have no doubt coded ways to automate this - but I have not as this would bypasses the intent of having the retry timer lengthen and simply make the server overload issues worse.

Normally the retry timer is pretty much latent as there are more WUs and more server capacity than these is compute resource/folders … at the moment with the rapid expansion of the folding community "the infrastructure" (hardware, software and tbh bloodware) cannot serve WUs to all the folders as soon as they need them … the team have been scaling up valiantly but for the moment there are times when folders just need to be patient - WUs will flow properly again as/when the infrastructure catches up.

It is a holiday weekend, people are in lockdown, everyone wants to help "now" … Just to manage expectations … It may not improve that much in the next few days - and even then it may takes weeks to really get on top of this.

I am not getting any WU as well. I just have installed F@H and I was wondering, how are my system's characteristics for tackling on the tasks and why I am cant get an assignment?

• :******************************* System ********************************
  CPU: Intel(R) Core(TM) i5-5200U CPU @ 2.20GHz
  CPU ID: GenuineIntel Family 6 Model 61 Stepping 4
  CPUs: 4
  Memory: 7.91GiB
  Free Memory: 4.55GiB
  Threads: WINDOWS_THREADS
  OS Version: 6.2
  Has Battery: true
  On Battery: true
  UTC Offset: -7
  PID: 8412
  CWD: C:\Users\Dick Thunderson\AppData\Roaming\FAHClient
  OS: Windows 10 Home
  OS Arch: AMD64
  GPUs: 0
  CUDA Device 0: Platform:0 Device:0 Bus:8 Slot:0 Compute:3.5 Driver:10.1
  OpenCL Device 0: Platform:0 Device:0 Bus:8 Slot:0 Compute:1.2 Driver:425.31
  OpenCL Device 1: Platform:1 Device:0 Bus:NA Slot:NA Compute:2.0 Driver:20.19
  Win32 Service: false

I have been getting this message quite a lot and not sure what it means but I quess the servers are full:

• 15:20:08:WU02:FS01:Connecting to 18.218.241.186:80
  15:20:08:WARNING:WU02:FS01:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
  15:20:08:ERROR:WU02:FS01:Exception: Could not get an assignment

Me too. No WUs available

17:06:02:ERROR:WU02:FS01:Exception: Failed to connect to 128.252.203.10:80: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond.
17:23:16:WU02:FS01:Connecting to 65.254.110.245:8080
17:23:16:WARNING:WU02:FS01:Failed to get assignment from '65.254.110.245:8080': No WUs available for this configuration
17:23:16:WU02:FS01:Connecting to 18.218.241.186:80
17:23:17:WARNING:WU02:FS01:Failed to get assignment from '18.218.241.186:80': No WUs available for this configuration
17:23:17:ERROR:WU02:FS01:Exception: Could not get an assignment

Also, I have 2 RX580 GPUs and they each take 4 hours to do a work unit. I leave it on 24/7, so my system alone hits your server to upload WU and download a new one 12 times a day.
Maybe make work units larger, to reduce the number of times we connect to your server, to reduce server load.

Under normal circumstances the retry timer is not used as the WUs throughput copes with demand … let the team catch up … server load isn't just about the number of connections - fewer larger WUs may mean fewer connections but would also mean more to download/upload the load may not improve - and if I understand it correctly the size of the WU is generally defined by the science research requirements.

Increasing the size of the WU would be in the time dimension, more steps. The download size would not change, but the upload size would increase as the amount of intermediate information included would be larger.

bloblobl0 wrote:I am not getting any WU as well. I just have installed F@H and I was wondering, how are my system's characteristics for tackling on the tasks and why I am cant get an assignment?...I have been getting this message quite a lot and not sure what it means but I quess the servers are full...

Welcome to the F@H Forum bloblobl0,

From the system configuration you posted, you only seems to have a CPU which is usable by F@H. AFAIK, there are plenty of CPU WUs available so it might take few attempts since the network and connection to the F@H Servers are the current bottlenecks. While things have improved from few weeks ago, there is more work to be done over the next coming weeks to hopefully address this situation.

Thank you PantherX, I am glad that I found out about this project and forum, glad I can help.

As I have seen the OS Arch is AMD64, so that's why my GPU can't be used for folding?

And I have successfully connected to some WU with my CPU, so it's fine. I can fold!

bloblobl0 wrote:Thank you PantherX, I am glad that I found out about this project and forum, glad I can help.

As I have seen the OS Arch is AMD64, so that's why my GPU can't be used for folding?

And I have successfully connected to some WU with my CPU, so it's fine. I can fold!

AMD64 is a cpu-architecture term. Both Intel, AMD and NVidia tech will work happily.