We need 360 more GPUs to get on top of the leaderboard
Now that Custom PC’s folding team is getting tantalisingly close to the number 1 spot, it’s becoming clear that our folding team has what it takes to be the top folding team in the world. Our production rate has already surpassed that of the teams at number 3 and 4 on the leaderboard, and we’ve worked out that we only need around another 360 GPUs folding in order to catch up with the production rate of the top team, [H]ardOCP.
As a result, we’re now on a mission to recruit those last few GPUs. We’ve just put up a new count-down thread on the folding forum to kick off the race to the top, so every time you add a GPU to our folding team, you can add your name to the thread and a new total of GPUs needed to take the number 1 spot. Okay, so it’s an unashamed rip-off of [H]ardOCP’s GPU countdown thread earlier this year, but [H]ardOCP’s GPU folding drive was hugely successful, and it would be amazing if we could repeat the same story on team Custom PC.
You don’t need a killer GPU to help out; it just needs to be compatible with Stanford’s latest GPU2 client. The client currently supports ATI Radeon GPUs from the 2000-series to the 4000-series, as well as any CUDA-supporting GeForce card – from the 8000-series and upwards. A 512MB GeForce 8800 GTS currently gets a PPD of around 4,500 with a quad-core CPU, so 360 of these would add another 1.62 million points a day to our team, taking us well over the production rate of the other top five teams.
So, if your GPU is currently sitting idle in 2D mode, then get it folding, add your name to the folding thread and make team Custom PC the number 1 folding team in the world. Not only will this be a tremendous success for the Custom PC community, but it will also give Stanford a lot more processing power.
Folding@home isn’t just about hardware bragging rights; it’s also a brilliant distributed computing project that researches protein folding. The project has already published a number of peer-reviewed results, and its aim is to gain a greater understanding of diseases such as Huntington’s and Alzheimer’s. Custom PC’s Folding@home team has been folding since 2004, and it’s currently at number 5 on the world folding leaderboard. You can join our team by using the team ID: 35947.
As soon as i get home ill stick my two BFG 280 OCX(overclocked a bit more). for the cause!!
Im running 8800GTX and e8400 and trying to get GPU and CPU clients to play nice. the CPU is running the console version and i set different machine IDs for each. Despite my manually (in task manager ) setting the affinity for the 2 processes (FAHCORE_1-gpu and FAH_82 - cpu) each time the console finishes a work unit it restarts the process and it ends back on the same core as the GPU - argh! any advice apart from sitting there and doing it manually...
For those who run Linux: http://gpu2.twomurs.com
How much will my electricity bill go up? 8800GTX SLi :)
> Lightning_Pete, you can try looking in the log file in the directory you installed the folding client in, it should be 'fahlog.txt' or similar.
9600GT 512MB or a PS3? Just out of interest?
If im actually folding?! i have the icon in the bottom right corner, but whenever i want to see the previewer for my 8800GTS 320mb folding it just crashes and the F@H closes..... How can i tell if im folding or not :/
Hey all, just started folding again, with 1 core of my 9800Gx2 and my Q6600 @ 3.6Ghz, so should get a health PPD :D FahMon is reporting about 6000ppd total from both :D
to everyone doing this. To be getting even close to the No1 spot is a fantastic acheivement. Especially at such a down and miserable time in the economy. This really is a noble cause and you all should be thanked for it.
http://www.facebook.com/group.php?gid=35882569440
i run a cpu client on my 8400 and a gpu client on my 8800gt. just be sure to leave one cpu core not folding if you run the gpu client, as the gpu runs about 30-40% of one cpu core. so a quad could run 3 cpu/ 1 gpu and still have plenty of grunt for surfing or other apps.
> Smiffy, the GPU client only uses a bit of CPU time so its easy to configure the two to run side by side with minimal reduction in output.
I just run the GPU client on its own, using cpu and gpu together just slows things down
That's no excuse! The more people using it the more work will be done to improve the service. Plus ATi have/are releasing a new driver that's supposed to sort out a lot of the problems with it. I'm gonna be pushing harder now and running F@H as much as possible!
hmm I would add my 4870, but it only gets like 4000ppd with all clocks at max... wish they'd sort the ati version :(
> Cerberus_xiii. I see you added your GPU in the forum thread so I guess you got it working alongside your PC. For anybody else wondering you need to set a separate machineid in the configuration for each folding client (i.e. each cpu and each gpu).
Quick question: is it safe to run a GPU console service at the same time as 2 CPU services?
I've noticed that its only the GPU2 client availabel now anyway, just got my 8800GT folding now. But, how do you run the GPU client and the CPU client at the same time, with the GPU client running, when I try to start the CPU client, it complains that another F@H process is runnning.
I only started folding last month and average around 5500ppd with a C2D E6600 and a 9800GTX+
Ia it the GPU2 client for single cards? I thought GPU2 was for things like 4870X2, 9800GX2 and so on.
Make a Comment
Fastest, cheapest 3G mobile broadband dongles from 3, Vodafone, T-Mobile and Orange
from just £10/month