ATI speeds to the front lines of health

Fight Alzheimer's disease with your GPU as Radeon boards take on more than gaming.

by

The video card isn't just for gaming anymore. Distributed computing pioneer Folding@Home announced today that it will release a client early next week that can take advantage of Radeon X1900 and X1950 class ATI GPUs to perform protein folding calculations on PCs. Folding@Home recently announced a client that will work on Sony’s PlayStation 3 console designed specifically to take advantage of the system’s multicore Cell processor.

Adding GPU processing power is expected to increase Folding@Home performance anywhere from 10 to 100 times compared to the current client that only uses the CPU, according to Folding@Home director and Stanford University associate professor of Chemistry Vijay Pande.

The Folding@Home scientific research project aims to increase our understanding of how proteins fold, with the hopes of applying the resulting work to cure diseases such as Alzheimer's, Parkinson's, and Huntington's. A protein consists of a long chain of amino acids that literally folds upon itself to perform vital functions. Properly folded proteins function within the body to enable many different kinds of processes. Genetic mutations can cause amino acid substitutions within a protein, which then cause it to fold improperly, rendering the protein either useless or actively harmful to the body.

Predicting how a protein folds requires immense amounts of computing power. Distributed computing gives scientists studying protein folding access to an enormous amount of idle computational power available within the general population. Each computer receives a tiny piece of the project to work on via the Internet. The PC sends the completed work back to Folding@Home servers and receives a new assignment.

According to Pande, Folding@Home currently has access to roughly 1.8 million CPUs, of which 200,000 are active, giving the group the equivalent of approximately 200 Teraflops of sustained computational power. In comparison, the largest supercomputer in the world, the US Department of Energy's Blue Gene supercomputer, has 131,000 processors and 280 teraflops (367 teraflops peak) of computing power. Now that Folding@Home has figured out how to use GPUs, Pande estimates that the power of its distributed network could rise to 1 to 10 petaflops.

While not as hardware intensive as common 3D benchmarking programs, Pande admits that running both your CPUs and GPUs at full throttle for extended periods can still generate a significant amount of heat. Folding@Home advises home client users to keep the computer case free of dust and to make sure all the fans are in working order.

The project expects to expand GPU support to the older Radeon X1800 series, but Nvidia GPU support is less certain due to "technical" difficulties.

Discussion

44 comments
exilim01
exilim01

Everything is about textures what about physics and game play ?

Shingo
Shingo

Its always good to see a uplifting news story in the midst of all of the depressing or crazy news in the world

_Sam_
_Sam_

can't the ps3 do that

SqueekyC1ean
SqueekyC1ean

i have fold@home on my pc but i use nVidia :(

Dantes_Monkey
Dantes_Monkey

isn't that the exact smae sort of thing the PS3 is gonna do?

chrisdojo
chrisdojo

guess AMD is making more $$$$....

foe666
foe666

"I wonder what ATI lovers that are also Sony haters think about this." Loving ATI and hating Sony would be stupid and pointless, since Sony is ATI's biggest customer...

fastpunk
fastpunk

Graphics cards have bigger processing power that CPUs, so the improvement will be substantial. It's nice to see commercial hardware used in other areas, like science. Good job ATI !

fissure_ocarina
fissure_ocarina

[This message was deleted at the request of a moderator or administrator]

rbarahona
rbarahona

Definitely this is simply awesome!!! To use all that kind of computer power in order to get knowledge of some process within the body is just great!!! This is just another great example of distributed tasks, just like the SETI project.

Sheiko
Sheiko

Too bad they don't help pay for electricity used when powering their calculations.

cjcr_alexandru
cjcr_alexandru

A good idea. If it doesn't burn my video card...

Kurushio
Kurushio

I think there are about a dozen other people that do thier own with the biggest being SETI, Folding, and AIDS i believe. When quad core comes out just imagine how much potential there would be with that combined with maybe quad SLI lol. I just hope the power company can keep up hehe.

TryMe01
TryMe01

wow, this is somethin gthat will out technology to good use, too bad somebody beat them to it...

Valen_Ca
Valen_Ca

I wonder how long before every other distributed computing project jumps on this.

Generic_Dude
Generic_Dude

Might not cure Alzheimer's, but running your CPU and GPU at full throttle does wonders for hypothermia. But, seriously, kudos to Folding@Home for fighting the good fight for people Forgetting@Home.

Kurushio
Kurushio

Yeah thats what i was kinda thinking of too. Even with my gpu at idle it is at 48C and my CPU at 100% for over 24hrs is just 49C. Still though if possible i would run it with gpu to get a speed boost for maybe an hour or so. Also it could be a great benchmarking test to make sure an overclock is stable lol.

chibi-acer
chibi-acer

Distributed computing will single-handedly bring about global warming. GPUs can sometimes be the hottest chip in the PC, I can't imagine running it 24/7.

666SolidSnake
666SolidSnake

Ati because x1950 just owned nvidia and the R600 will own nvidia's new card as well. by by nvidia

EMAN2g
EMAN2g

Nvidia for PCs and ATI for consoles all I gotta say...

Betasheet
Betasheet

Vicred, Yeah, i actually thought from reading the description that it was to solve the structure, not look at folding dynamics. I do work in a lab (structural biology). I still have hesitations, as targeting folding proteins with drug treatments seems tricky at best, as you have to consider spatial and temporal limitations (getting the drug to the protein is fine, but getting it to interact only within a specific timeframe may be difficult). In my experience, pure bioinformatic data has been inaccurate most of time, but like I said, that has been my experience. Anyways, good luck with your research, and we will see what this folding@home produces.

vicred
vicred

Betasheet, you miss completely the point. The objective of folding simulation is not to replace the NMR/X-ray structure, but to be able to study the little difference in the structure of aprotein when you introduce a few (1-3) mutations or how can interact with other molecules (medicament, for examples) and if this interaction can alter its structure and in what way. To do this calculation you always need a template and this is provided by the NMR/X-ray structure. We are really far away to be able to simulate the folding of a protein without any template, and this project can help to do a step more in that way, but the important thing for structure simulation of protein is not to replace the current methods of determine structure, but to provide us with the structure of slightly different proteins and how this protein can interactuate with other molecules. By the way the simulations are not so bad as you say, and more often than not they are quite similar to the experimental data. Im my laboratory (yes I work in a lab, and I presume you too) we use the simulations to estimate what residues mutate in order to obtain better mutants proteins (biological engineering), usually to increase stability. And you know what, they say me mutate this to this and about 50% of the time they are true! This can seem very little but is a lot of more than do several azar mutations and expect for a lucky shot.

Betasheet
Betasheet

X-Ray Crystallography rarely uses antibodies for protein crystallization. As for difference between NMR and X-Ray, they are some difference between structures, but the RMSD is still extremely high in 99% of the cases. The majority of proteins they have modelled using energy mimization algorithms have not matched the calculated structure to a significant RMSD. NMR and crystal data has also been shown to be correct in some studies utlizing small angle X-Ray and neutron scattering experiments. This folding@home is more about sparking public interest than it is actually creating biologically relevant data. EDIT Anyways stus, maybe this will produce something interesting, maybe it won't. At the very least it will generate interest in biochemistry and protein folding, which isn't a bad thing at all.

stus
stus

"This approach of determining protein structures simply by minimizing the energy terms sounds great in theory, but in practice the computed structure rarely matches the structure solved by X-ray crystallography or NMR. Even when the program calculates a fold, it would have to be checked by another method before anything biologically relevant could be ascertained, as there is no way of telling if the structure is correct or not. This whole project is just a waste of bandwidth in my scientific opinion. " NMR and X-ray crystallography are so much more reliable? They don't fully match each other most of the time, with NMR structures bieng very flexible at edges of the protein. The NMR structure and X-ray crystallography techniques study proteins in different states for starters, NMR is in solution and X-ray crystallography studies static crystals often containing antibodies. So who's to really say which technique is correct? But at least this is finally a use for my degree in biochemistry on a game site.

silverstar1982
silverstar1982

it'd be nice if they released it for lower power X1x-based cards. Put my X1600 Pro to use doing something useful while it's idle, right alongside my CPU doing it's own thing with folding.

SilV3RSix
SilV3RSix

Deimos27 and Pablo620 ... and at the end of the day I'll be dling this program the same as you guys. Lighten up. It's a joke. Besides, as my previous post shows, I took the time to do some research into the network and see what it's all about, so cast your stones somewhere else.

Betasheet
Betasheet

This approach of determining protein structures simply by minimizing the energy terms sounds great in theory, but in practice the computed structure rarely matches the structure solved by X-ray crystallography or NMR. Even when the program calculates a fold, it would have to be checked by another method before anything biologically relevant could be ascertained, as there is no way of telling if the structure is correct or not. This whole project is just a waste of bandwidth in my scientific opinion.

Deimos27
Deimos27

SilV3RSix "Maybe they'll learn to treat Alzheimers'.... I hope they can treat Alzheimers'... Maybe they can treat... what? huh? Alzheimers'?" Get out of that couch... Put away that controller... Get a life... Theres something called REALITY out there bro... Its nice to know that companies are contributing something...ANYTHING... in the war against terminal illnesses. Hats off to these companies...

luckjon
luckjon

Now if there was a patch for the nVidia dual core GPUs....

Pablo620
Pablo620

I wonder what ATI lovers that are also Sony haters think about this. They were probably bashing it on PS3 and now that something similar is being done else where will they still be angry about it. I think its great to help cure these things. My grandfather had this before it died and it isnt a joke. IM LOOKING AT YOU SilV3Rsix you douche bag

Kurushio
Kurushio

Ive been folding for them for about a year now (on gamespots team who's id to join is 32721) and although it does use 100% of cpu power i dont notice anything that really slows down with just surfing the net or chatting on an IM. If i play a game though i of course turn it off but otherwise it runs from startup to shut down. If they ever figure out how to make use of the 7900GT then ill be glad to put that to a little bit of work as well (though probably not all day long since i would like that to last longer than a year lol).

the_real_VIP
the_real_VIP

[This message was deleted at the request of a moderator or administrator]

SilV3RSix
SilV3RSix

Maybe they'll learn to treat Alzheimers'.... I hope they can treat Alzheimers'... Maybe they can treat... what? huh? Alzheimers'?

PlanoPunisher
PlanoPunisher

Yeah, I won't be using this. My PC is hot enough already.

bored88
bored88

They should make a F@H app for the psp. Maybe then they would get used!

UnknownPerson55
UnknownPerson55

its always nice to see something like this being done.

Amidamaru_01
Amidamaru_01

That's a great alternative use for such powerful hardware.

blackIceJoe
blackIceJoe

Wow so this is the stuff that will be in the PS3 sounds pretty neat. The Folding@Home program is a really weird thing but if it can help people I am all for it. I hope Radeon X1900 and X1950 class ATI GPUs perform really well for people doing this on the PC because if it works out it could really help out a lot of people.