GPU2 configuration guide

[Sunin]

I know for a fact that this is still hit or miss. I tried it on another machine and it worked without the command... regardless if you do have trouble you may wanna look at the command if your not having any issues fold on!

 

[Dan-515]

The guide is very good. I just had to follow it as I had some huge Windows failure requiring a format. I thought I would be able to configure the clients back up but it was to complicated to remember

It's strange I re-installed XP almost on auto-pilot as I've done it so many times, only when I got to the desktop did I realise I had been using Vista and what I'd done

 

[gwarren007]

I was curious about something and I think it's probably best to ask here. What are the minimum CPU requirements for GPU folding? Can any CPU be effective as long as it is able to feed the GPU, or are there specific generations that demarcate which CPUs can and cannot fold with the GPU client?

I have a Celeron 430 (11.40 off eBay) with 1gb ddr2-mvr (7.98 of ebay) on a ib9 40?(thanks EA). The ram is set to ddr2-667, then I overclock the fsb till I got ddr2-800 ram speed. Now its at 2.17 and currently putting out 4100ppd with a 9600gt (99.00 circut city). 650/1799/900

I had no different ppd when using a 2140 with NV affinity.

 

[Pooch666]

I wish mine would run at 1750 shaders
ATI Tool starts reporting errors at 1500 and system crashes come shortly thereafter. By contrast I have my 8800GS at 1500 shaders 100% stable and I think it will go faster. WTF!! Sorry, I guess this isn't really an overclocking thread, but if anyone has any insight...

You should be getting more out of your 88gs shaders. I run mine at 1850 and it's been completely stable for weeks.

 

[chrisf6969]

Now, if Nvidia would just release an AGP 8800GT I'd buy about 10 of them for all the boxen that I have laying around & really crank out some PPD. (I have a lot of 2-3Ghz P4's at work & various places that I could recruit.)

 

[Sunin]

I wish mine would run at 1750 shaders
ATI Tool starts reporting errors at 1500 and system crashes come shortly thereafter. By contrast I have my 8800GS at 1500 shaders 100% stable and I think it will go faster. WTF!! Sorry, I guess this isn't really an overclocking thread, but if anyone has any insight...

You should be getting more out of your 88gs shaders. I run mine at 1850 and it's been completely stable for weeks.

That is of course UNLINKed right... because if your boosting your gpu along with your shader it is goign to be limited alot earlier. you must unlink the two and ramp the shaders.

 

[APOLLO]

Now, if Nvidia would just release an AGP 8800GT I'd buy about 10 of them for all the boxen that I have laying around & really crank out some PPD.

Agreed, but highly doubtful. AGP seems to be Ati's domain.

 

[JaYp146]

Has anyone tried to run the Folding client on their 8400 mobile GPUs? Was thinking about folding on my Thinkpad but wanted to see if anyone else had success ...

 

[chrisf6969]

Has anyone tried to run the Folding client on their 8400 mobile GPUs? Was thinking about folding on my Thinkpad but wanted to see if anyone else had success ...

I've run it on an 8300GS, which is a crappy OEM desktop card. Not sure how it exactly compares to an 8400 mobile.

It only gets about 350PPD. Or one 480 point WU every 33hrs or so.

 

[Supercharged_Z06]

Awesome guide guys - thanks everyone for the excellent info and tips! Finally have both of my 8800GTX's folding together now for Team [H]... major PDD boost! WOOT!!

 

[Sunin]

Grats, was there anything that wasn't crystal clear? We are always trying to improve it

 

[Servant of Shodan]

I've got what may be a very simple or very difficult question, but it has certainly stumped me thus far.

I have two 7800GTX cards in SLI - these will never be able to run CUDA, therefore, no F@H GPU2 client for them.

However, I wanted to have a second monitor setup (nvidia does not let you use a second monitor connected to the SLI setup ), and I specifically wanted to run F@H on it. Now, I've got two PCIe x16 slots already populated, but have extra PCIe x1 slot. The only CUDA-enabled card that has a PCIe x1 interface in production anywhere is the Quadro NVS 290 x1. I scored one off eBay for $50. Sweet!

I've got the correct CUDA driver enabled and my Quadro NVS 290 and the 2 7800GTX's work just fine for all the regular Windows stuff. I downloaded and installed the F@H GPU2 client (the new non-beta version), and when I tried to run it, it said, "Your video card is not supported..." or something synonymous...

I've uninstalled, reinstalled, taken cards out, put them in different orders, etc. I cannot get F@H to recognize the NVS 290, but is is absolutely shown at nvidia's site as working with CUDA - it is a G86-based card. Does anybody know perhaps how I can force F@H to run on this card? I thought about using the force to GPU 1, 2, 3, etc switch, but how do I know what GPU number F@H considers the NVS290 to be?

Thanks for any help.

 

[Sunin]

http://www.flickr.com/photos/24842648@N02/sets/72157604195510800/

Last 3 pics are of the vga dummy plugs I made the other day Sometimes pictures help you see what you need to do.

I used 68 to 100 watt resistors they all seemed to work ok... I had a hard time finding 75ohm 1/4 watt resistors.