Nvidia And Stanford Finalizing Folding@Home Client For GeForce GPUs

[Mr.Guvernment]

ouch!

http://www.tomshardware.com/news/Nvidia-Stanford-Folding-Home,5654.html

Santa Clara (CA) - During Nvidia Editor’s Day, we learned that Nvidia and the Folding@Home research group led by Vijay Pande are making final preparation to launch the first version of the Folding@Home client for Nvidia graphics processors.

The unveiling of the client is set for the next week as part of the launch of Nvidia’s GT200 GPU series. Owning such a card will have its benefits in Folding@Home and will outrun Radeon 3870 cards. The new GeForce cards are expected to hit more than 650 nanoseconds of protein simulation in a single day, while the Radeon HD 3870 is stuck at about 170 ns. The Playstation 3 is able to produce "only" 100 ns of simulation, while a quad-core CPU creates an output of just four nanoseconds. For those who are keeping count: The GeForce GPU will be about 163 times faster than a quad-core processor in this specific application.

Nvidia founded Team "Whoopass", which consists only of several computers that are running the Folding@Home GPU client. Even with just 4-5 test machines, the team quickly moved into the top 5% of all contributors by sheer processing power. Dr. Vijay told us that if only 1% of all CUDA-capable users would start using Folding@Home in their spare time, the Folding@Home machine would quickly be considered the fastest performing HPC computer in the whole world - hitting about 60-80 Peta FLOPS of processing power.

Folding@Home for Nvidia CUDA-capable graphics cards (GeForce 8 and above) should become available next week. The codename for this client is GPU2/NVIDIA. The GPU1 client was retired, while GPU2 client will continue to be updated for both Nvidia and ATI cards.

ATI was first with a client for the Folding@Home project, which was released back in September of 2006 for the X1900 series of cards. Back then, the cards topped out at 375 GFlops. The next GPU generation should provide more than double the horsepower.

 

[Automata]

...so they are going to leave the points like that? That doesn't seem very fair.

I hope they also mentioned that the processor is extremely versatile and can do calculations that the GPU isn't capable of.

 

[dfonda]

I have to question whether any GPU is built to run 24/7.

If it gets 3k ppd and I'm in.( Not that I care about points. lol)

 

[Shiggity]

Nice find Guvernment!

@PPD -
GPU2 PPD increase

With new revisions to the GPU2 core code, the GPU2 code is doing more science per ever in the same amount of time. In the spirit of having the points equate to how much science is being done, we are going to increase the PPD of the GPU2 benchmark machine (ATI 3850 board -- see the FAQ for details) to 1500 PPD. We will revalue the most recent WU's (472X) and all future WU's will use this new PPD for benchmarking.



They will likely be increasing all GPU PPD's in the near future. The amount of science they can get from a GPU is pretty sick and I'm sure they really want it. Vijay knows how to get it done

Let's find some cures!

 

[ihrsetrdr]

I have to question whether any GPU is built to run 24/7.

If it gets 3k ppd and I'm in.( Not that I care about points. lol)

Yea, I'll have to see how these pan out; I got too crazy too early on the first GPU bandwagon.

 

[harlam357]

I hope they also mentioned that the processor is extremely versatile and can do calculations that the GPU isn't capable of.

Of course they didn't... it's an nVidia marketing parade! They would never admit that their GPUs are inferior, in any manner, to Intel or the CPU in general.

I think it's funny how Intel refuses to sell nVidia a QuickPath license... haha Mr. SLI hog, who holds the keys to the bus now?

...and to be on-topic... I'll only look into nVidia (or any GPU) Folding if it's economically sensible - PPD/watt. I am also curious to see how this current (HD 2xxx/3xxx & GeForce 8xxx/9xxx) generation of graphics cards is going to hold up in the long-term, 24x7 usage category.

Of course I know we all want to see HD 4xxx and GT 2xx numbers too.

 

[veryhumid]

how about sli?

edit:

What about multi-gpu support?

We hope to have multi-gpu support in the future.

 

[Automata]

how about QUAD sli?

Fixed!

 

[veryhumid]

lol, thanks!

If it supports it and the points are good I will splurge for another gx2

edit: zomg it's in the FAQ: http://folding.stanford.edu/English/FAQ-NVIDIA

 

[dz_jad]

http://folding.extremeoverclocking.com/team_summary.php?s=&t=131015

 

[dz_jad]

http://fah-web.stanford.edu/cgi-bin/main.py?qtype=teampage&teamnum=131015

 

[Shiggity]

Holy shnazz @ that team and PPD.

I wonder if 1 280GTX = 6 active processors? Or if he has 6 graphics cards. (Tony_Stark)

 

[dz_jad]

I wonder what cards will be usable. This poor college student might find a couple of cheap used ones and SLI them and take advantage of this mobo (see sig)

 

[Shiggity]

The G80 cards and later for Nvidia. (8800GTX, 8800ULTRA, 8800GT, 8800GTS, yeah anything with an 8 and higher )

 

[dz_jad]

maybe when I graduate in a year I can buy me a nice little GPU folding rig... We shall see.
It all depends on the CPU impact on the performance. If I can get by with a nice little single core feeding it on a cheap (yet good) mobo, get a couple rigs up and going and get back in the deal.

 

[Shiggity]

It's looking like a 280GTX is up there with a Q6600 folding rig in PPD? Except the 280GTX costs more than the entire Q6600 folding rig

What I want to know is what's the cheapest CPU you can buy to get the full PPD / Power out of an overclocked 280GTX, (overclocked CPU of course). Or am I mis-understanding this, does the GPU not rely on the CPU at all to do these science calculations?

 

[dz_jad]

what can a quad crank out now-a-days? 'cause raising the PPD for GPU to 1500 for the ATI cards...and the NVIDIA being able to blow that out of the water....Its hard to believe.

**hard to believe that it could barely keep up with a quad...

 

[dz_jad]

does the GPU not rely on the CPU at all to do these science calculations?

The GPU needs a CPU to "feed it"
With the more powerful GPU2 client and GPUs, they need a pretty dang good processor dedicated to it...
GPU folding takes 1 GPU and 1 CPU per GPU client

 

[Shiggity]

I believe a 3.6GHz Q6600 gets like 3400-4000PPD.

Now they said the PS3 is ~100ns and the 280GTX was doing ~700ns, and a PS3 folding nonstop gets like ~900PPD. So 6300 PPD for a 280GTX, more if they are increasing GPU2 point WU's.

I still don't know if it'll be worth it in the wattage / electricity bill because you'll need a highly overclocked CPU to keep up with the already huge power consumption of that card. We'll have to see when some of our folders get it and the real PPD #'s surface.

It really depends how much they inflate the PPDs of the GPU WU's.

 

[Mr.Guvernment]

Apparently not ALL the members on that team are using GPU only, others are using CPU, they are giong to update it apparently so you can tell who is using GPU and who is CPU.

still impressive.. i dont like the point change, give GPU more points, guess cause they can complete faster but ya... GPU 24/7 full load?