When setting up the client, what arguments do I need to put in? And secondly, where do I put in the arguments? Shortcut - Properties - End of Taget Name? I asked this in another thread, but, seeing as I'm here, with running 2 x285s in SLI, do I still need to run 2 separate clients, or just one? Do I still need a dummy plug for the card? What can I use as a dummy plug for the card? Thanks!
You put the arguments just as you stated, make a shortcut for the client and add the arguments at end of target name. You don't say what clients your going to be running so I'll assume your just going for GPU atm. I can't remember whether Stanford added SLI support, so go with disabling SLI for folding. 2x 285s will get you sickass points if you set it up right. What OS are you running? I'm using W7 64-bit (which every sane person should be using) and I folded on 3 GPUs in one rig, with only 1 GPU connected to a monitor, and I didn't need any dummy plugs or anything like that. 1) Head to http://folding.stanford.edu/English/DownloadWinOther. You want Windows Vista/7 console client which is at the bottom of the GPU section, the first one above the SMP section. 2) Extract the ZIP into two different folders. IE. C:\First GTX and C:\Second GTX. This is so you can run two clients simultaneously. 3) Create two shortcuts, one for each client, and move them to your desktop or wherever you like. Add the argument "-configonly" onto each. This is where you can configure the client. 4) Make sure you put in your folding name, AlchemyFire, and the team, 35947. You can ignore passkey. You can just press enter to skip through the other options, but make sure you stop and change the MachineID. The MachineID helps identify the clients when your running multiple clients on one rig, so make sure the values are different for each client. 5) Now you have done that, remove the "-configonly" argument on the end of the shortcut. Add "-gpu 0 -forcegpu nvidia_g80" on the end of your first 285 client - this will run the client on the 285 which is outputting to your screen. Add "-gpu 1 -forcegpu nvidia_g80" on the end of your second 285 client. 6) Download HFM.NET (http://code.google.com/p/hfm-net/) and Hardware Monitor (http://www.cpuid.com/hwmonitor.php). HFM monitors your clients for you, just install the program and add the directories of the two clients into it and you will get a PPD figure. Hardware Monitor helps you monitor temps of your graphics cards. 7) Start up the clients. They should start folding. You should see the temperature on BOTH your cards rise quickly. After the clients have done a few % of their workunits, try refreshing the information in HFM. You should the info turn green and give a PPD figure. You should get around 8K for each client - if your getting only 8K in total, check your temperatures. If only one of your GPUs is getting hot, then you folding on one card and you've configured it wrongly. Good luck!
Thanks Ph4ZeD! Unfortunately only running Vista 64bit atm Was debating whether to run the GPU and the SMP client - have an i7 950 in the box as well
That CPU will give you 10K PPD at stock mate, so you could be looking at 26K PPD with no overclock, which if you kept that up would nearly get you in the top 20 producers
GPUs typically run hotter than CPUs. Around 80-90C at max load is fine, and you will get slightly higher temps due to the fact you have the cards next to each other.
Yes. You don't need to run multiple clients even on quad/hexa core CPUs. Its only when you have multiple GPUs that you need more clients.
What we do on Bit-Tech is when someone has been a big help, we click the +rep button under their name /apologies for the shameless rep plug ><
Am I doing something wrong here: I downloaded the SMP MPICH client and the new drop in binary. I installed the client, dropped in the binary, ran the install.bat file, ran the client using -configonly argument, set up everything including bigadv passkey, changed the client argument to -smp and ran. All I keep getting is: + Attempting to get work packet Passkey found - Connecting to assignment server + No appropriate work server was available; will try again in a bit. +Couldn't get work instructions - Attempt # to get work failed, and no other work to do. Waiting before retry. And this just carries on. Am I doing something incorrectly? Edit: Changed the -smp argument to -smp 8, still nothing
