GPU won't fold

[PantherX]

The log in the first post suggests a Whitelist error:
0x10de:0x1049:2:1:GF116 [GeForce GT 620]

1 -> Pre-Fermi GPUs
2 -> Fermi GPUs
3 -> Kelper GPUs

Since you have a Fermi, the whitelist is incorrect so I will report it and hopefully, it will be updated.

[doublestorm]

[Napoleon]

Looks like PantherX spotted the root cause of the problem already. I forget, is ticket #883 the proper place to follow up on the progress?

@OP: could you try updating the whitelist manually, just in case? viewtopic.php?f=83&t=21205 for instructions. Probably won't solve the problem until the list is updated, but it can't hurt to try.

FWIW, I suppose this GPU is a rebranded Fermi since GPU-Z says GF119 instead of GK???. Pretty obvious from the clocks too, shaders are at 2x instead of 1x.

[PantherX]

Looks like PantherX spotted the root cause of the problem already. I forget, is ticket #883 the proper place to follow up on the progress?...

That ticket is fixed and closed. I did inform the appropriate personal and the changes are in the pipeline, might take sometime to be pushed out.

...@OP: could you try updating the whitelist manually, just in case? viewtopic.php?f=83&t=21205 for instructions. Probably won't solve the problem until the list is updated, but it can't hurt to try...

No need to try that since even if you do it, nothing will change since the whitelist has to be updated on the Servers so it won't do any good.

...FWIW, I suppose this GPU is a rebranded Fermi since GPU-Z says GF119 instead of GK???. Pretty obvious from the clocks too, shaders are at 2x instead of 1x.

That is correct. Looks like Nvidia might want to consider a better naming scheme based on the architecture and not the next number in the series:
http://www.anandtech.com/show/5845/nvid ... nto-retail
http://www.engadget.com/2012/05/20/nvid ... no-kepler/

[doublestorm]

Gents, Thanks so much for looking at this for me. One last question on this is how will I know when it is updated on the servers, and what steps do I need to take on my end after it is? I am assuming I will have to download the new GPUs.txt and replace it in the FAHClient folder? Will I need to reinstall or just shut down and restart?

[7im]

This has been updated in the source GPUs.txt. I don't know how long it takes to get it published. Download the new file when it becomes available.

Also, you may have to remove and readd that GPU slot if the change is not recognized automatically.

[PantherX]

If you want to help test a use case for us, please read the green text and skip the blue text. If you are not interested in testing a use case, please skip the green text and read the blue text. Black text can be read in either case. Thanks.


Before you update your GPUs.txt file, you add the GPU Slot to your FAHClient and make sure that it download a FahCore_11 WU and let is "run" (it won't progress or anything). You can than exit the FAHClient, delete the GPUs.txt file and when you restart the FAHClient, it will automatically download the updated GPUs.txt file and what happens next is important:
1) Verify that the GPU is correctly detected as Fermi
2) Look if the GPU Slot is still trying to process the FahCore_11 WU or will it automatically dump it.
A) If it automatically dump the WU, then that it perfect
B) If it doesn't try to dump the WU, than an enhancement bug can be written for this behavior.
Do make sure that you post the log files showing all the above steps and outcome if possible



For you to update the GPUs.txt file, here is what you do:
1) Remove the GPU Slot and save the changes
2) Exit the FAHClient
3) Go to the C:/Users/Tom/AppData/Roaming/FAHClient/ folder and delete the GPUs.txt file
4) Restart your FAHClient and it should automatically download the latest GPUs.txt file
5) Add a GPU Slot and enjoy folding on it


No need to reinstall the software.

As 7im stated, we are unsure of when it will be pushed. If you want, here is what you can do:
1) Open this link in the browser -> https://fah-web.stanford.edu/file-relea ... c/GPUs.txt
2) Search for 0x10de:0x1049 and depending on what you see, do this:
A) Nothing if you see this -> 0x10de:0x1049:2:1:GF116 [GeForce GT 620]
B) Update the GPU.txt file if you see this -> 0x10de:0x1049:2:2:GF116 [GeForce GT 620]

[doublestorm]

Yes, I can absolutely do that to help out. Will let you know how it goes after it gets added.

[PantherX]

Thanks for helping us out

BTW, the GPUs.txt file has been updated with your GPU Model so you can experiment with your FAHClient whenever possible.

[doublestorm]

Yes, I already checked it....saw that it was there and went through the steps. It DID in fact identify it as a Fermi and downloaded an 0x15. I'm just semi-patiently waiting to see if it starts folding.

[doublestorm]

Success!!! Thank you so much for your assistance with this!

[PantherX]

Glad that it worked out for you.

Before you downloaded the GPUs.txt, was the GPU Slot already having FahCore_11 WU or not? If it had the FahCore_11 WU, after the GPUs.txt file was updated, it did automatically get FahCore_15 WU or not?

If you are unsure of the answer, please post the log files, current and previous ones (present in the log folder) for us to view.

Thanks.

[doublestorm]

Yes, when I added the GPU slot in, it downloaded a FahCore_11 and it was "running"
I exited the client, deleted the GPUs.txt file and restarted the client.
Verified that the GPU was detected as a Fermi
Checked the WU it was working on and at that time I saw it was 0x15

The core_11 was replaced by the 15 as you said it would above.