• Log in with Facebook Log in with Twitter Log In with Google      Sign In    
  • Create Account
  LongeCity
              Advocacy & Research for Unlimited Lifespans

Photo
- - - - -

PNY Folding@Home Distributed Folding Competition


  • Please log in to reply
18 replies to this topic

#1 maestro949

  • Guest
  • 2,350 posts
  • 4
  • Location:Rhode Island, USA

Posted 11 December 2008 - 06:51 PM


Our folding@home prize now has competition! The graphics card maker is following our lead and providing incentive through prizes. It would be great to see this become a trend.

Attached Thumbnails

  • pny2.jpg

Edited by maestro949, 11 December 2008 - 06:51 PM.


#2 chungenhung

  • Guest, F@H
  • 114 posts
  • 20

Posted 11 December 2008 - 06:56 PM

Our folding@home prize now has competition! The graphics card maker is following our lead and providing incentive through prizes. It would be great to see this become a trend.

eVGA also encourage their users to fold.
Those companies are using this F@H to improve their public image.

#3 maestro949

  • Topic Starter
  • Guest
  • 2,350 posts
  • 4
  • Location:Rhode Island, USA

Posted 11 December 2008 - 07:33 PM

Our folding@home prize now has competition! The graphics card maker is following our lead and providing incentive through prizes. It would be great to see this become a trend.

eVGA also encourage their users to fold.
Those companies are using this F@H to improve their public image.


Which might backfire considering the carbon footprint these puppies leave!

#4 Mind

  • Life Member, Moderator, Secretary
  • 17,116 posts
  • 2,009
  • Location:Wausau, WI

Posted 12 December 2008 - 05:23 PM

A good use of energy, I think. It can be justified if it helps cure aging and saves lives. Go PNY and eVGA!

#5 poser

  • Guest, F@H
  • 116 posts
  • 4

Posted 12 December 2008 - 06:32 PM

Carbon footprint... is it my fault that coal plants are still used instead of nuclear power?

#6 kismet

  • Guest
  • 2,984 posts
  • 424
  • Location:Austria, Vienna

Posted 13 December 2008 - 01:11 PM

Our folding@home prize now has competition! The graphics card maker is following our lead and providing incentive through prizes. It would be great to see this become a trend.

eVGA also encourage their users to fold.
Those companies are using this F@H to improve their public image.


Which might backfire considering the carbon footprint these puppies leave!

You know what I keep on saying on the issue? We could be buying or renting Bluegene/P or similar clusters for that money to get the best carbon footprint... or even better wait 10 years until the same performance is available at 1/100th of the carbon footprint. Or we could fold, now! There's always a slightly better option around the corner when it comes to hardware. And anything that we can do now involves certain trade-offs (renting a supercomputer is not easy nor will it give us easy PR that folding@home can gives us, nor would it enable a pretty neat and friendly competition).
The only issue worth mentioning (for imminst users) is IMHO: What's better, paying energy costs for folding vs investing the money in SENS? Both? What percentage to spend on each?

Edited by kismet, 13 December 2008 - 01:11 PM.


#7 maestro949

  • Topic Starter
  • Guest
  • 2,350 posts
  • 4
  • Location:Rhode Island, USA

Posted 13 December 2008 - 06:40 PM

You know what I keep on saying on the issue? We could be buying or renting Bluegene/P or similar clusters for that money to get the best carbon footprint... or even better wait 10 years until the same performance is available at 1/100th of the carbon footprint. Or we could fold, now! There's always a slightly better option around the corner when it comes to hardware. And anything that we can do now involves certain trade-offs (renting a supercomputer is not easy nor will it give us easy PR that folding@home can gives us, nor would it enable a pretty neat and friendly competition).
The only issue worth mentioning (for imminst users) is IMHO: What's better, paying energy costs for folding vs investing the money in SENS? Both? What percentage to spend on each?


An interesting debate.

The problem with big-ass iron like Bluegene/P computers is how quickly they become obsolete. By the time you get them up and running, build the software specifically for their architecture, the newer models are already far superior in performance. They also cost a fortune in manpower to setup and maintain and have a large footprint. They are a single point of failure too from both a project risk perspective as well as putting too many eggs-in-one-basket perspective.

A distributed computing model may burn more human effort and suck the electrical grids dry as they continue to expand, but can evolve much more organically and leverage the power of large numbers. The distribution of hardware and software installations is saving Stanford Folding and other DC projects enormous amounts of capital that can instead be poured into the science.

A point I raised in another thread is that many people are reluctant to turnover their hard-earned cash for the sake of pooling it for big projects. Look at the fierce resistance to pay taxes as simple example. Another would be the explosion in family-based philanthropic endeavors rather than just turning the money over to central planners. Sure teamwork, divide and conquer and pooling resources sounds like a great idea but it's doomed to failure as the the bureaucracy that usually emerges, swallows it up while new smaller efforts emerge around them and bypass them. There are trust issues and an inherent belief by many that they can do better by themselves. I suspect that the most successful projects of the future will utilize these selfish tendencies in an organic and evolutionary manner with limited centralization.

The key is rearchitecting projects to leverage these concepts. F@H is one of the best examples we have under our meme's umbrella but it's bottom up science and alone, has a long-term ROI inflection point. SENS is a top down model, also with a similar inflection point. Where the two meet in the middle, i.e. the seemingly impenetrable complexity barrier to both their success is where our longevity singularity will emerge. Thus I would reprhase your question as to:

How do we invest in expediting the slow collision course these two approaches are on?

My Opinion? Invest in a middle out strategy as well. i.e. systems biology where we build bigger and better tools to aggregate the wealth of biological data into computationally accessible means for the sake of building the models that bridge the gap, which I believe, will finally allow us to engineer cost-effective aging treatments.

Edited by maestro949, 13 December 2008 - 06:51 PM.


#8 poser

  • Guest, F@H
  • 116 posts
  • 4

Posted 16 December 2008 - 02:52 PM

Bluegene is expensive, about 5.5 times that of PC solution. It consumes about 40% of the electricity as equivelant amount of PC's if my calculation were right. This difference isn't big enough to make a good solution.

#9 kismet

  • Guest
  • 2,984 posts
  • 424
  • Location:Austria, Vienna

Posted 16 December 2008 - 03:32 PM

Bluegene is expensive, about 5.5 times that of PC solution. It consumes about 40% of the electricity as equivelant amount of PC's if my calculation were right. This difference isn't big enough to make a good solution.

I made a very simple calculation some time ago max. FLOPS/power draw of the best bluegene/P system back then, versus FLOPS/estimated power draw of FAH. I think the difference was one or two orders of magnitude, but the calculations were probably flawed.
My calculations looked something like this FWIW:
560.000 participants, ~1000TFLOPS, computer average draw 200Watt, overall 1200MW
BlueGene/P (357MFLOPS/W) needs 2,8MW for a similar performance.

When I divide 151/357 (PS3/bluegene MFLOPS/W) I also get ~42%, but I am not sure if any CPUs were as efficient as a PS3 one year ago and there are probably thousands of old pentiums and similar CPUs, which have a dismal performance/W, still contributing to FAH. I believe the combined efficiency of FAH is way below that of an average PS3.

The key is rearchitecting projects to leverage these concepts. F@H is one of the best examples we have under our meme's umbrella but it's bottom up science and alone, has a long-term ROI inflection point. SENS is a top down model, also with a similar inflection point. Where the two meet in the middle, i.e. the seemingly impenetrable complexity barrier to both their success is where our longevity singularity will emerge. Thus I would reprhase your question as to:

How do we invest in expediting the slow collision course these two approaches are on? 

My Opinion? Invest in a middle out strategy as well. i.e. systems biology where we build bigger and better tools to aggregate the wealth of biological data into computationally accessible means for the sake of building the models that bridge the gap, which I believe, will finally allow us to engineer cost-effective aging treatments.

Nice catch. Interesting explanation of those project involving a top-down respectively bottom-up approach. So those two are basically independent and synergistic?

Edited by kismet, 16 December 2008 - 03:34 PM.


#10 maestro949

  • Topic Starter
  • Guest
  • 2,350 posts
  • 4
  • Location:Rhode Island, USA

Posted 22 December 2008 - 02:24 AM

Here is the prize list of stuff they are giving out for the above contest. You can stay on our team and also participate in this contest. Let's show our support of the contest by joining. Hopefully it will lead to more!



Graphics Cards
Model Part # Retail Value Qty
F@H 9800 GTX + 512MB VCG98GTXPXPB $229.99 2
9800 GT 1GB VCG981024GXPB $199.99 2
9600 GSO 768MB OC VCG96GSO7GXPB-OC-BB $149.99 4

Memory
Model Part # Retail Value Qty
2GB 800/4 Kit MD2048KD2-800-X4 $44.99 4
4GB 800/4 Kit MD4096KD2-800-X4 $69.99 4

Peripherals
Model Part # Retail Value Qty
4GB Optima Attache P-FD4GBATT2-FS $19.99 4
8GB Optima Attache P-FD8GBATT2-FS $29.99 8
16GB Attache P-FD16GATT2-FS $39.99 4

#11 MichaelGR

  • Guest, F@H
  • 45 posts
  • 0
  • Location:Near Ottawa

Posted 22 December 2008 - 03:23 AM

Carbon footprint... is it my fault that coal plants are still used instead of nuclear power?


In most places, you can sign up for "green power" with your utility so that your dollars go toward wind/solar/hydro/whatever instead of more polluting alternatives.

#12 maestro949

  • Topic Starter
  • Guest
  • 2,350 posts
  • 4
  • Location:Rhode Island, USA

Posted 27 February 2009 - 10:20 PM

They finally got this contest up and running. It starts March 4th. I'll be trying for a spot in the top 10. Wish me luck :)

http://portal.xlr8ga...ge=folding-info

To sign up you'll need a passkey. Go here to get one:

http://fah-web.stanf...n/getpasskey.py

#13 Live Forever

  • Guest Recorder
  • 7,475 posts
  • 9
  • Location:Atlanta, GA USA

Posted 28 February 2009 - 12:00 AM

They finally got this contest up and running. It starts March 4th. I'll be trying for a spot in the top 10. Wish me luck :)

Hopefully that does not mean you will be leaving our team.

#14 maestro949

  • Topic Starter
  • Guest
  • 2,350 posts
  • 4
  • Location:Rhode Island, USA

Posted 28 February 2009 - 12:49 AM

Nope. Don't need to. I'm here for eternity :)

#15 Live Forever

  • Guest Recorder
  • 7,475 posts
  • 9
  • Location:Atlanta, GA USA

Posted 28 February 2009 - 12:54 AM

Nope. Don't need to. I'm here for eternity :)

You better be. I'll be checking back every 1,000 years or so to make sure.

#16 dnamechanic

  • Life Member
  • 1,518 posts
  • 0
  • Location:Dallas, TX

Posted 28 February 2009 - 01:52 AM

... I'll be trying for a spot in the top 10. Wish me luck :)

Yes, here's wishing you good luck and most importantly 'good folding'.

#17 Heliotrope

  • Guest
  • 1,145 posts
  • 0

Posted 02 May 2009 - 06:39 PM

hey maestro made to top 10?

#18 maestro949

  • Topic Starter
  • Guest
  • 2,350 posts
  • 4
  • Location:Rhode Island, USA

Posted 02 May 2009 - 11:18 PM

Not sure. They had problems out of the gate with the data transfers from F@H to their db so the stats were all fubar. I stopped paying attention after that :-D. I did win a random prize though (flash drive)!

#19 Heliotrope

  • Guest
  • 1,145 posts
  • 0

Posted 03 May 2009 - 03:05 AM

Not sure. They had problems out of the gate with the data transfers from F@H to their db so the stats were all fubar. I stopped paying attention after that :-D . I did win a random prize though (flash drive)!



nice! too bad their competition is only a 1 month push. Even tho their folding prize stopped, the enthusiasm still keeps them pushing by top 200 team




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users