Can't get both GPUs to Fold

Q

qdemn7

Supreme [H]ardness
Joined
May 2, 2002
Messages
4,531
I have a problem with getting both of my graphics card to fold. The GTX260 runs great, and I’m doing about 7k PPD. But for some reason I just cannot get the 8800GT to start running. I’ve disabled the 800GT as a Physx card. The GTX 260 is in PCI Express slot 1 nearest the N/B, the 8800Gt is in the middel PCI Express slot. Both at X16 slots. PCI Express slot 3 is empty. I’ve done as the sticky states to do to set up two shortcuts, but every time I try to run two clients, they both go to the GTX260. I’m stumped, any advice appreciated.

Here’s my system:
Intel Core 2 Quad Q9450 @ 2.66 ghz (stock)
DFI X38LT LP LT
BFG GTX 260 (216 cores / 590 core clock)
BFG 8800 GT (112 cores / 626 core clock)
8192 MB of DDR2 PC2-6400
Thermaltake Toughpower 1200
Thermalright Ultra 120 eXtreme
Dell 2405FPW monitor
OCZ Vertex 60 GB SSD
Samsung 1TB hard drive
Seagate 1 TB hard drive
Windows 7 Pro 64 bit
Nvidia 6.23 F@H GPU client (systray)

 
How are your flags set up? They should be like this: -gpu x -forcegpu nvidia_g80, where the "x" is 0 or 1 depending on which card the client is for. Also, what driver version are you using?
 
-gpu 0 -forcegpu -nvidia_g80 -verbosity 9
-gpu 1 -forcegpu -nvidia_g80 -verbosity 9

Driver version 196.21 (64 bit)
 
There should be no dash before nvidia_g80. It should be "-forcegpu nvidia_g80".
 
Is there a monitor or dummy plug on the 2nd card?

Side note... I thought folding with 2 different chipset cards really messed things up and killed ppd of the more powerful card...??
 
Ok, thanks, fixed that.

Here's what I keep getting from the 8800GT - UNSTABLE MACHINE

[00:34:45] + Attempting to send results [February 27 00:34:45 UTC]
[00:34:45] - Reading file work/wuresults_09.dat from core
[00:34:45] (Read 2357 bytes from disk)
[00:34:45] Connecting to http://171.64.65.71:8080/
[00:34:46] Posted data.
[00:34:46] Initial: 0000; - Uploaded at ~3 kB/s
[00:34:46] - Averaged speed for that direction ~11 kB/s
[00:34:46] + Results successfully sent
[00:34:46] Thank you for your contribution to Folding@Home.
[00:34:50] Trying to send all finished work units
[00:34:50] + No unsent completed units remaining.
[00:34:50] - Preparing to get new work unit...
[00:34:50] + Attempting to get work packet
[00:34:50] - Will indicate memory of 8190 MB
[00:34:50] - Connecting to assignment server
[00:34:50] Connecting to http://assign-GPU.stanford.edu:8080/
[00:34:50] Posted data.
[00:34:50] Initial: 43AB; - Successful: assigned to (171.67.108.11).
[00:34:50] + News From Folding@Home: Welcome to Folding@Home
[00:34:50] Loaded queue successfully.
[00:34:50] Connecting to http://171.67.108.11:8080/
[00:34:51] Posted data.
[00:34:51] Initial: 0000; - Receiving payload (expected size: 47226)
[00:34:52] - Downloaded at ~46 kB/s
[00:34:52] - Averaged speed for that direction ~83 kB/s
[00:34:52] + Received work.
[00:34:52] Trying to send all finished work units
[00:34:52] + No unsent completed units remaining.
[00:34:52] + Closed connections
[00:34:57]
[00:34:57] + Processing work unit
[00:34:57] Core required: FahCore_11.exe
[00:34:57] Core found.
[00:34:57] Working on queue slot 00 [February 27 00:34:57 UTC]
[00:34:57] + Working ...
[00:34:57] - Calling '.\FahCore_11.exe -dir work/ -suffix 00 -nocpulock -checkpoint 15 -verbose -lifeline 4352 -version 623'

[00:34:57]
[00:34:57] *------------------------------*
[00:34:57] Folding@Home GPU Core
[00:34:57] Version 1.31 (Tue Sep 15 10:57:42 PDT 2009)
[00:34:57]
[00:34:57] Compiler : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86
[00:34:57] Build host: amoeba
[00:34:57] Board Type: Nvidia
[00:34:57] Core :
[00:34:57] Preparing to commence simulation
[00:34:57] - Looking at optimizations...
[00:34:57] DeleteFrameFiles: successfully deleted file=work/wudata_00.ckp
[00:34:57] - Created dyn
[00:34:57] - Files status OK
[00:34:57] - Expanded 46714 -> 252912 (decompressed 541.4 percent)
[00:34:57] Called DecompressByteArray: compressed_data_size=46714 data_size=252912, decompressed_data_size=252912 diff=0
[00:34:57] - Digital signature verified
[00:34:57]
[00:34:57] Project: 5766 (Run 6, Clone 124, Gen 1744)
[00:34:57]
[00:34:57] Assembly optimizations on if available.
[00:34:57] Entering M.D.
[00:35:03] Tpr hash work/wudata_00.tpr: 3689336029 3153036780 3561954513 2740745329 3374077584
[00:35:03]
[00:35:03] Calling fah_main args: 14 usage=100
[00:35:03]
[00:35:03] mdrun_gpu returned
[00:35:03] Going to send back what have done -- stepsTotalG=0
[00:35:03] Work fraction=0.0000 steps=0.
[00:35:07] logfile size=4948 infoLength=4948 edr=0 trr=25
[00:35:07] + Opened results file
[00:35:07] - Writing 5486 bytes of core data to disk...
[00:35:07] Done: 4974 -> 1869 (compressed to 37.5 percent)
[00:35:07] ... Done.
[00:35:07] DeleteFrameFiles: successfully deleted file=work/wudata_00.ckp
[00:35:07]
[00:35:07] Folding@home Core Shutdown: UNSTABLE_MACHINE
[00:35:11] CoreStatus = 7A (122)
[00:35:11] Sending work to server
[00:35:11] Project: 5766 (Run 6, Clone 124, Gen 1744)
[00:35:11] - Read packet limit of 540015616... Set to 524286976.


[00:35:11] + Attempting to send results [February 27 00:35:11 UTC]
[00:35:11] - Reading file work/wuresults_00.dat from core
[00:35:11] (Read 2381 bytes from disk)
[00:35:11] Connecting to http://171.67.108.11:8080/
[00:35:11] Posted data.
[00:35:11] Initial: 0000; Conversation time very short, giving reduced weight in bandwidth avg
[00:35:11] - Uploaded at ~6 kB/s
[00:35:11] - Averaged speed for that direction ~10 kB/s
[00:35:11] + Results successfully sent
[00:35:11] Thank you for your contribution to Folding@Home.
Click to expand...
 
Intelman34 said:
Is there a monitor or dummy plug on the 2nd card?

Side note... I thought folding with 2 different chipset cards really messed things up and killed ppd of the more powerful card...??
Click to expand...

No monitor, and what do you mean by "dummy plug"?
 
qdemn7 said:
Wow, thanks for that. I did not know that, If you have a second monitor hooked up to the other card, does it have to be on?
Click to expand...

I leave my monitor off on the second card with no probs.
 
DooKey said:
I leave my monitor off on the second card with no probs.
Click to expand...
Thanks to all of you.

Zero82z for the flag issue.
Intelman34 for the monitor
DooKey for your advice.

Turns out the the "missing" monitor was the problem. I dug out my old CRT monitor and hooked it up, rebooted the system, and now both cards are running fine.
 
Last edited:
Good news indeed. I need to get one of my 275s reinstalled. Now if I had something that could fit in that third slot....
 
You must log in or register to reply here.
Back
Top