Linux GPU - No Folding Was: GTX 680 recognized as Fermi?

[Dark_n_Beyond]

I'm trying to get my GTX 680 classified to fold, but not having much luck. I did notice that the log gives this card a GF114 id, which I thought was fermi. I was expecting to see GK114. Anyway, all I get is

Code: Select all

01:49:12:WARNING:WU00:FS01:Failed to get assignment from 'assign-GPU.stanford.edu:8080': Empty work server assignment
01:49:12:ERROR:WU00:FS01:Exception: Could not get an assignment.

I assume this is because of the incorrect id?

Code: Select all

*********************** Log Started 2012-11-21T01:01:05Z ***********************
01:01:05:************************* Folding@home Client *************************
01:01:05:    Website: http://folding.stanford.edu/
01:01:05:  Copyright: (c) 2009-2012 Stanford University
01:01:05:     Author: Joseph Coffland <joseph@cauldrondevelopment.com>
01:01:05:       Args: --child --lifeline 959 /etc/fahclient/config.xml --run-as fahclient
01:01:05:             --pid-file=/var/run/fahclient.pid --daemon
01:01:05:     Config: /etc/fahclient/config.xml
01:01:05:******************************** Build ********************************
01:01:05:    Version: 7.2.9
01:01:05:       Date: Oct 3 2012
01:01:05:       Time: 10:34:06
01:01:05:    SVN Rev: 3578
01:01:05:     Branch: fah/trunk/client
01:01:05:   Compiler: GNU 4.4.7
01:01:05:    Options: -std=gnu++98 -O3 -funroll-loops -mfpmath=sse -ffast-math
01:01:05:             -fno-unsafe-math-optimizations -msse2
01:01:05:   Platform: linux2 3.2.0-1-amd64
01:01:05:       Bits: 64
01:01:05:       Mode: Release
01:01:05:******************************* System ********************************
01:01:05:        CPU: Intel(R) Core(TM) i5-3570K CPU @ 3.40GHz
01:01:05:     CPU ID: GenuineIntel Family 6 Model 58 Stepping 9
01:01:05:       CPUs: 4
01:01:05:     Memory: 7.75GiB
01:01:05:Free Memory: 7.19GiB
01:01:05:    Threads: POSIX_THREADS
01:01:05: On Battery: false
01:01:05: UTC offset: -6
01:01:05:        PID: 1024
01:01:05:        CWD: /var/lib/fahclient
01:01:05:         OS: Linux 3.2.0-23-generic x86_64
01:01:05:    OS Arch: AMD64
01:01:05:       GPUs: 1
01:01:05:      GPU 0: NVIDIA:2 GF114 [GeForce GTX 680]
01:01:05:       CUDA: 3.0
01:01:05:CUDA Driver: 5000

[7im]

Hello Dark_n_Beyond, welcome to the forum.

First, exit fahcontrol, and then delete the GPUs.txt file from the FAHClient Data folder.

Restart FAHControl, and it should download an updated file, containing this updated entry:

0x10de:0x1180:2:3:GK104 [GeForce GTX 680]

If 0x1180 is not the device ID for this GPU, please use GPUZ to find the ID and post it so we can get the list updated as needed. Thanks.

[bruce]

You need to update your GPUs list. Upgrade to V7.2.x which should update it automatically or follow the manual procedure described in the topic above.

You didn't post enough of the log to see which version of the FahCore is being used. That's currently the only difference and it's still working it's way through the distribution channels. After that change is complete, Development will go to work on the next generation of FahCores so there will be more updates when they are ready for distribution.

[Dark_n_Beyond]

Thanks for the fast replies.

I wonder if the source of my problem is running linux? I am running version 7.2.9, which was installed 2 days ago, I think. Linux is new to me, so I'm still learning the ins and outs, but I can't find a GPUs list. I did find what I believe to be a GPU Id though

Model: GeForce GTX 680
IRQ: 16
GPU UUID: GPU-8d3d438c-3b79-42fc-7907-d120869cb489
Video BIOS: 80.04.47.00.80
Bus Type: PCI-E
DMA Size: 40 bits
DMA Mask: 0xffffffffff
Bus Location: 0000:01.00.0

Perhaps I should see if I can get it working in windows first.

[Joe_H]

Yes, that is part of your problem. GPU folding is currently not supported on Linux systems. There is a third party hack to run folding with some nVidia GPU's on Linux using WINE, but that only works using older cards and with the V6 client.

[Dark_n_Beyond]

Thanks Joe...I guess that explains that. I must have missed that somewhere along the line. Thanks for your help!