It is currently Sun Jul 13, 2014 3:21 pm

All times are UTC - 8 hours




Post new topic Reply to topic  [ 44 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: This is a sad thread. The High Cost of Folding.
PostPosted: Tue Sep 25, 2007 12:24 pm 
Offline
*Lifetime Patron*

Joined: Fri Apr 04, 2003 6:21 pm
Posts: 4247
Location: Undisclosed but sober in US
http://www.ocforums.com/showthread.php?p=5247792

Yeah costs for FAH can become a bummer with a good size farm, but why should we foot the entire bill ourselves? I wrote this on their forum:

Why shouldn't Folding be tax deductable if you meet all the rules?

This would be all I can think of:
1. The IRS recognizes FAH as a legitimate charity.
2. FAH thanks you for your donation of xxxxx.
3. You have a separate electrical meter, which is connected only to your Folding PCs. You need paper proof of your costs, not just the reading from a [$20 plug in] wattage meter. Too bad this necessitates the monthly flat fees as well.

This should work like people who donate their time and car for Meals on Wheels, they get to write off xx cents per mile.

Also see if you can fold strictly off peak at a reduced rate.
Aris

_________________
People who put money and political ideology ahead of truth and ethics are neither patriots nor human beings.


Top
 Profile  
 
 Post subject:
PostPosted: Tue Sep 25, 2007 6:54 pm 
Offline

Joined: Thu Oct 21, 2004 4:49 pm
Posts: 336
Location: Upper left hand corner, USA
Why bother with the electric meter?
Since contributions to folding are already measured in terms of
computations performed - they could just offer a receipt based on a standard cost per point value. [Based on efficient current technology - taking into account amortized value of computer plus typical electricity rate.]
Like the standard milage rate - you can keep track of your actual expenses and deduct them if you do it enough to make it worth while, or just use the standard rate.
Figuring the standard rate would require a bunch of assumptions, but given all the server farms out there, people must already have done this sort of calculation - so it shouldn't be hard to come up with a cost that is reasonable and defensible.
A standard rate/point value would make the deduction a lot more accessible to people. It would probably make folding more attractive. (It might also make people more aware of the environmental costs of their folding. Which might not make them fold as much, but might help on the global warming and energy conservation fronts. I think the bit in the folding FAQ about power usage is far to glib and underplays the costs involved.)

Using a standard rate would also help limit abuses by people running other programs on the folding machines, etc.

Since I read the thread about Points per day per watt I also wondered about why they don't offer donation receipts for taxes.
I also wondered about how the IRS and others view this activity with regards to the finances and tax returns of the Folding project.
I am not an accountant, but as far as I gather, it can be important for non-profits to track and report the value of contributions (either in terms of taxes, or in terms of applying for grants, etc.) Of course if they aren't a non-profit, then it is even more important that they account for the contributions.


Top
 Profile  
 
 Post subject:
PostPosted: Tue Sep 25, 2007 7:24 pm 
Offline

Joined: Thu Oct 21, 2004 4:49 pm
Posts: 336
Location: Upper left hand corner, USA
Did a little searching on the topic - thought this message might
be interesting/useful for those wishing to pursue the matter.

http://einstein.phys.uwm.edu/forum_thre ... true#42900

(Relates to a different project - but the principle is probably the same.)


Top
 Profile  
 
 Post subject:
PostPosted: Tue Sep 25, 2007 8:16 pm 
Offline
Friend of SPCR

Joined: Sun Jun 03, 2007 12:03 am
Posts: 777
Location: Norway
My old man commented on it, thankfully he doesnt care too much, atleast not yet.. might have to shut off one of my C2D's atleast 50% of the time.. But now with the "cold season" arriving, it does help keep the house warm, especially here with 6" insulation in all the walls.

Edit: we're using about 50kWh/day on average, thats about a 2kW load 24/7... I wonder how much of that is from the PCs..
(we're got one p3 gw, one c2d ws, one c2d f@h box, one fileserver, my sisters 2500+ and three laptops)

_________________
Workstation | HTPC | 9.1TB | 19.1TB


Top
 Profile  
 
 Post subject:
PostPosted: Wed Sep 26, 2007 5:33 am 
Offline
*Lifetime Patron*

Joined: Fri Apr 04, 2003 6:21 pm
Posts: 4247
Location: Undisclosed but sober in US
scdr wrote:
Did a little searching on the topic - thought this message might
be interesting/useful for those wishing to pursue the matter.

http://einstein.phys.uwm.edu/forum_thre ... true#42900

(Relates to a different project - but the principle is probably the same.)


Interesting thread. Like the others, I don't even want to try write off anything other then the actual electricity used, and honestly, my $20 power meter could accurately measure exactly that.

Of course if you have a 200 pc folding farm that requires a/c as well that might be a harder sell. 8)

Wibla wrote:
might have to shut off one of my C2D's atleast 50% of the time

Watch those deadlines. :shock:

_________________
People who put money and political ideology ahead of truth and ethics are neither patriots nor human beings.


Top
 Profile  
 
 Post subject:
PostPosted: Sun Sep 30, 2007 10:47 pm 
Offline

Joined: Sun Sep 30, 2007 8:41 pm
Posts: 35
I got my Q6600 rig.. ran BOINC 100% for a month about 18/7.. (ie, 100% when not in use)... bil for that month doubled.

BOINC went bye bye.... Q6600 is too much of a pig for that.


Top
 Profile  
 
 Post subject:
PostPosted: Mon Oct 01, 2007 12:13 am 
Offline

Joined: Wed Jun 02, 2004 10:31 am
Posts: 520
Location: Las Cruces, NM
Folding at Home makes no sense when viewed from a global benefit economic perspective. So far the results haven't made a significant impact medically and combined people are using literally 1000s of extra kilowatts to run folding at home. I can see running it on an already running computer, but running machines just for the purpose of running folding at home in my mind is just wasted power. If they were really making progress and curing cancer I would agree, but with the project now years long with little more than a bit of understanding gained I'm prone to disagree that such a project deserves any kind of tax boost.


Top
 Profile  
 
 Post subject:
PostPosted: Mon Oct 01, 2007 9:20 am 
Offline
*Lifetime Patron*

Joined: Fri Apr 04, 2003 6:21 pm
Posts: 4247
Location: Undisclosed but sober in US
It's not fair to address waste this way. I feel the entire line of Pentium 4's and D's are a waste of electricity.

There are other people who have SLI setups, the fastest most power hungry processors, a pair of G8800 NVidia cars (at 200 watts each), with a 1000 watt non-APFC power supply playing games on-line and using Internet bandwidth with a total disregard.

But you did bring up an interesting point that many may have missed. Many SPRC staff members have completely stopped folding, even during times they could running it as a service when they use their PCs for work. Why is that? I could agree with minimizing costs, that makes sense, but if you're doing something on a pc already then running full tilt won't be a big deal.

On the other hand check out the costs related to a full day in a nursing home for someone who's completely gone because of Alzheimers or dementia. And since in many instances the government is going to pick up that tab, well, all of us will be paying for it anyways.

As for research being useful, I decide to look at this way. I've seen people "researchers" caught wasting funds. Here I know where all my funds are going, and none of my expenditures are on such things as refinishing antique pianos like I've seen on "20/20" years ago.

_________________
People who put money and political ideology ahead of truth and ethics are neither patriots nor human beings.


Top
 Profile  
 
 Post subject:
PostPosted: Tue Oct 02, 2007 1:45 pm 
Offline

Joined: Sat Jun 07, 2003 8:41 am
Posts: 33
aristide1 wrote:
On the other hand check out the costs related to a full day in a nursing home for someone who's completely gone because of Alzheimers or dementia. And since in many instances the government is going to pick up that tab, well, all of us will be paying for it anyways.


As you said this is a very cost effective dontation.

The non-economic side of the equation is that the patient you are helping to save might be YOU.


Top
 Profile  
 
 Post subject:
PostPosted: Tue Oct 02, 2007 2:00 pm 
Offline

Joined: Wed Jun 30, 2004 10:54 pm
Posts: 273
This is scary. Considering I am going to get my q6600 to run seti 24/7. How much power can the q6600 use.
Thanks


Top
 Profile  
 
 Post subject:
PostPosted: Tue Oct 02, 2007 2:30 pm 
Offline
Patron of SPCR

Joined: Mon May 17, 2004 1:59 pm
Posts: 117
Location: Manchester, UK
Elixer wrote:
Folding at Home makes no sense when viewed from a global benefit economic perspective. So far the results haven't made a significant impact medically and combined people are using literally 1000s of extra kilowatts to run folding at home. I can see running it on an already running computer, but running machines just for the purpose of running folding at home in my mind is just wasted power. If they were really making progress and curing cancer I would agree, but with the project now years long with little more than a bit of understanding gained I'm prone to disagree that such a project deserves any kind of tax boost.


I would agree with Elixer that Folding at Home is wasteful. If you go to the Folding at Home website and look at their published papers, there is very little to suggest anything that is going to benefit mankind directly in the near or not so near future. Set against this the monetary cost to individuals and the huge environmental cost.

I say stop folding and turn your computers off when they're not in use.


Top
 Profile  
 
 Post subject:
PostPosted: Tue Oct 02, 2007 2:37 pm 
Offline

Joined: Tue Sep 20, 2005 6:55 am
Posts: 5085
Location: UK
Quote:
I say stop folding and turn your computers off when they're not in use.


and donate the money you would have spent Folding to some cancer or Alzheimer's research charity.

ps. did you know researchers now think Alzheimer's might be a type of diabetes?

Quote:
Considering I am going to get my q6600 to run seti 24/7.


SETI is even more of a waste of time and money than Folding.


Top
 Profile  
 
 Post subject:
PostPosted: Tue Oct 02, 2007 6:05 pm 
Offline
*Lifetime Patron*

Joined: Fri Apr 04, 2003 6:21 pm
Posts: 4247
Location: Undisclosed but sober in US
I'm not one of those lunatics spending $500/month on electricity, but I do acknowledge that if I lose my brains all the money I've saved won't me a thing to me.

I like the idea that I am still on control. That's more important than all the money.

Another question - How does FAHs contribution towards Alzheimers compare to other types?

_________________
People who put money and political ideology ahead of truth and ethics are neither patriots nor human beings.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Oct 03, 2007 6:54 pm 
Offline
*Lifetime Patron*

Joined: Fri Apr 04, 2003 6:21 pm
Posts: 4247
Location: Undisclosed but sober in US
jaganath wrote:
ps. did you know researchers now think Alzheimer's might be a type of diabetes?


You know older people are often border-line with diabetes II, so instead of drugs they are put on a more strict diet and monitored so they don't get worse. I know someone who went years like that. Of course his sugar wasn't normal normal, just barely in the acceptable range.

Doctors now suspect that may be a mistake, and are starting to suggest drug treatment right away in such a situation, that harm occurs anyways when the numbers are a bit off.

One could argue that's drug makers talking, but what if it's true?

_________________
People who put money and political ideology ahead of truth and ethics are neither patriots nor human beings.


Top
 Profile  
 
 Post subject:
PostPosted: Thu Oct 04, 2007 1:54 am 
Offline

Joined: Sun Oct 29, 2006 2:46 pm
Posts: 10
The FAH project is a basic (or strategic) research program. The objective is to gain knowledge on the behaviour of key components of the human body on a molecular level. The results will most likely contribute to the basis of future applied medical research on cures and prevention of cancer and other diseases, within a 10-30 years perspective. This is how it works.

No research projects come with a guarantee of success or useful results. However, even negative and disappoining results may be important for further research. Most of the applied cancer research produce negative results, nonetheless is it necessary to search every corner until we know they are empty. Keep this in mind, and be proud of your contribution when folding.


Top
 Profile  
 
 Post subject:
PostPosted: Thu Oct 11, 2007 2:03 pm 
Offline

Joined: Sun Oct 09, 2005 8:35 am
Posts: 1253
Location: Pleasanton, CA
Rather than F@H, I'd recommend World Community Grid. It's kind of the mutual fund of volunteer computing, running 5 or so different projects at once. One of them is of course Human Proteome Folding. They also provide periodic updates on the results of the projects.

BTW, they recently switched to BOINC, which can run multicore CPUs effectively, unlike the previous UD agent. In addition to being able to run all my cores at a base load that is just below where my CPU fan ramps up, my daily scores went up over 4x by switching to BOINC.

As for energy consumption, I am spending about 75 cents/day at top PG&E (California) rates doing this.

_________________
i7 2600K CPU@4.4 GHz, Asrock Z68, 8GB Corsair Vengeance 1866 CL9, Intel 335 240GB SSD + Samsung HD502HI 500GB, Internal i7 graphics, Antec P180 case, Seasonic X-400 fanless PS, Megahalems CPU HS, Nexus 3-pin & AC PWM fans ~ 600 RPM, AcoustiPack foam, homemade ducts.


Top
 Profile  
 
 Post subject:
PostPosted: Sun Oct 14, 2007 1:50 pm 
Offline

Joined: Thu Oct 21, 2004 4:49 pm
Posts: 336
Location: Upper left hand corner, USA
Elixer wrote:
Folding at Home makes no sense when viewed from a global benefit economic perspective. So far the results haven't made a significant impact medically and combined people are using literally 1000s of extra kilowatts to run folding at home.


Bit of an under-estimate, quick calculation puts electrical donations to folding something on the order of
10,000,000 kilowatt-hours
(worth about $2million at current prices, and added something like 10,000 metric tons of extra CO2 in the air)

(per my post in http://www.silentpcreview.com/forums/vi ... sc&start=0 )

Elixer wrote:
I can see running it on an already running computer

Since going from idle to fully loaded for a running computer can significantly raise the electricity usage
(albeit not as much as going from off to fully loaded), even this decision isn't clear-cut.

Elixer wrote:
I'm prone to disagree that such a project deserves any kind of tax boost.


I don't think non-profit status is based on any particular determination of social benefit/worth. Or if it is, it is prety broad.
While I don't think folding has a particularly good return on investment, I am sure there are lots of tax-deductible organizations that I would consider far less worthy.
So it seems reasonable that those donating should receive deductions.

I also suspect that as people become aware of the impact of what they are doing, they may make wiser decisions.
So if it is brought to their attention that folding 24/7 is costing them something like $130/year (per folding website),
they might decide to do something else with those resources, or they might decide it was money well spent.
Either way it is enough of an investment that it is worth thinking about.


Top
 Profile  
 
 Post subject:
PostPosted: Sun Oct 21, 2007 8:40 am 
Offline

Joined: Thu May 19, 2005 12:48 pm
Posts: 191
Location: Shi-Khan: Vulcan or MosEisley Tattonnie
Instead of beefing about F@H energy demand, why not start looking at all the parasitic loads in your house? You'd be surprised at how much they can add up to because they're hidden.

Parasitic loads include your TV (instant on - keeps picture tube warm), your Stereo (no tubes, why does it need standby power?) all the damn digital clocks in your house (I prefer windup as power is unreliable during winter), VCR's, Cable Boxes and a whole rash of other stuff.

As an example: A 5watt 24/7 parasitic load adds 3.6KW per 30 days. It's a simply calculation 5x24x30 = 3600 watts or 3.6KW per month. What percentage of your monthly power usage is that?

Now if you want a horrifying number, multiplay that 3.6KW parasitic load against every house in your neighborhood. Scarry isn't it and if everyone of us kills such loads, what would our national power consumption do? In my neck of the woods, we'd get rid of the need for 2 future Nuclear plants, 4 coal fired plants and the associated Carbon footprint from them. Doesn't this make far more sense to you?

So get off your butt and beef to your governmental enviromental agency about all the parasitic loads and get them to crack down on devices continuing to use as much as 1/4 their full power demands when shutdown because when shutdown, it shouldn't use any power should it? even when it's got a remote. I know I can certainly stand to wait a half second for my stereo to power up. Hell it takes a couple of seconds for the lead in on a CD to give you any music since the record studio already took that into account.

_________________
AsRock Q87M vPro - Intel E3-1230V3 Xeon - 4x Kingston KVR13N98S/4 (4GB)
Silverstone SST RL01B-USB 3.0 - Coolermaster Hyper T4 - Asus DVD Burner
Crucial M500 120GB SSD (boot) Samsung 1TB Spinpoint - WD 2TB Green
Seagate Barracuda 3TB - EVGA GT640 w/2GB Video Card (dual Slot)
Locked Multiplier @ 10x


Top
 Profile  
 
 Post subject:
PostPosted: Wed Nov 07, 2007 12:26 pm 
Offline
*Lifetime Patron*

Joined: Fri Apr 04, 2003 6:21 pm
Posts: 4247
Location: Undisclosed but sober in US
Fastturtle has a point. I'd like to add that most new power supplies have the 5V standby now boosted to a max of 3 amps, which means the thing can draw as much as 15watts turned off if it was 100% efficient, which it isn't.

The other day I unplugged the TV and DVD player in the house. They were off when I plugged them in again, but the spark and pop made it clear they still wanted juice.

------------------------

To return to the original post the guy who was bailing out because of $500 electric bills decided to dump his 3 Pentium D's, which I thought would giving him the maximum reduction in electricity while reducing his points the least. I think he's still folding with 3 duals and a quad. With some other stuff I believe, his cost reduction was $200 for the month of October.

------------------------

On a more strange note there are new soap dispensers at work. They obviously run on electricity. Uh, did I miss something? Was here something wrong with the old ones? If you're worried about spreading germs you need to address the people who don't wash their hands at all. What's next? A toilet that needs electricity to flush?

_________________
People who put money and political ideology ahead of truth and ethics are neither patriots nor human beings.


Top
 Profile  
 
 Post subject:
PostPosted: Wed Nov 07, 2007 3:22 pm 
Offline
Friend of SPCR

Joined: Sun Jun 03, 2007 12:03 am
Posts: 777
Location: Norway
We physically disconnect power from the TV stuff in the livingroom in the evening, and I always shut off amplifiers and monitors via the power button before I go to bed.. I leave the pcs running tho, atleast most of them.. folding goes first, haha. Joke aside tho, we usually have one in-floor heating thingie going during the fall and winter, but now with two new machines running 24/7 we havent turned them on this year.. sick :)

_________________
Workstation | HTPC | 9.1TB | 19.1TB


Top
 Profile  
 
 Post subject:
PostPosted: Mon Nov 19, 2007 10:26 am 
Offline
*Lifetime Patron*

Joined: Fri Apr 04, 2003 6:21 pm
Posts: 4247
Location: Undisclosed but sober in US
Pity, scroll down to the power/watts graph, Phenom does not do well.

http://enthusiast.hardocp.com/article.h ... lzdWFzdA==

_________________
People who put money and political ideology ahead of truth and ethics are neither patriots nor human beings.


Top
 Profile  
 
 Post subject:
PostPosted: Mon Nov 19, 2007 12:13 pm 
Offline
*Lifetime Patron*

Joined: Fri Apr 04, 2003 6:21 pm
Posts: 4247
Location: Undisclosed but sober in US
cmthomson wrote:
Rather than F@H, I'd recommend World Community Grid. It's kind of the mutual fund of volunteer computing, running 5 or so different projects at once. One of them is of course Human Proteome Folding. They also provide periodic updates on the results of the projects...
.

The Proteome project has been marked complete.

http://www.worldcommunitygrid.org/stat/ ... e=proteome

_________________
People who put money and political ideology ahead of truth and ethics are neither patriots nor human beings.


Top
 Profile  
 
 Post subject:
PostPosted: Sun Nov 25, 2007 3:53 pm 
Offline

Joined: Sun Nov 25, 2007 3:18 pm
Posts: 1
Elixer wrote:
Folding at Home makes no sense when viewed from a global benefit economic perspective. So far the results haven't made a significant impact medically and combined people are using literally 1000s of extra kilowatts to run folding at home. I can see running it on an already running computer, but running machines just for the purpose of running folding at home in my mind is just wasted power. If they were really making progress and curing cancer I would agree, but with the project now years long with little more than a bit of understanding gained I'm prone to disagree that such a project deserves any kind of tax boost.



There is a lot of confused thinking about the environmental costs of Folding at Home. Whether folding generates carbon dioxide depends on the geographic location of the computer and the outside temperature…. Sounds stupid but bear with me.

For example in most parts of Canada supplemental heat is required in most homes and offices for about 8 months of the year. In most homes this heating is provided by burning of fossil fuels of some type while in a small minority it is provided by resistance heating using electricity. The main reason why resistance heating is not more popular stems from the price per unit of heat ( BTU, calorie, joule, kw/h etc) and historically electricity has been quite a bit more expensive than fossil fuels. A computer is essentially a resistance heater. Eventually 100% of the energy used by a computer ends up as heat in the home or office. If the home or office requires supplemental heat it will need less of it due to the heat from the folding computer. The thermostat will switch on less often and the requirement for supplemental heat will be less with a corresponding reduction in the use of natural gas, oil or what ever. Whether folding increases carbon will depend on what heat source it is substituting for and how clean the electricity is. If for example you live in Quebec or Manitoba or British Columbia where electricity comes almost exclusively from hydroelectricity and your supplemental heat in your home is a fossil fuel then almost certainly you are actually helping the environment by folding as there is less net carbon entering the atmosphere because you have essentially substituted clean electricity for a dirty fossil fuel to heat your home. If you come from Ontario the situation is essentially neutral ( carbon emission factor .223 grams CO2equivalent / kwh) with electricity contributing as much carbon to the atmosphere as does burning fossil fuel in your home. In many states especially those that rely heavily on coal fired electricity generation the carbon emission factors are much higher such that folding increases carbon output throughout the year.

The situation in the summer months is different in that folding generates heat when it is not wanted and additional energy is expended getting rid of the heat with an air conditioner. Thus folding in months when supplemental heat is not required does increase carbon in the atmosphere in almost all locations except those with very green electricity and even in those areas it would be preferable to export that electricity to substitute for dirty electricity elsewhere.

Thus the short version is that folding is bad for the environment in the summer. In the winter is acts as a heat source and if your electricity source is cleaner then your usual supplemental heat source then folding will actually decrease atmospheric carbon dioxide output.

Whether folding is good for your wallet is a whole different story. In most states and provinces the cost of electricity per unit of heat if a lot more than the cost of natural gas or oil per unit of heat so you will be losing money by substituting electrical heat from folding from your current heat source. Scientific advancement usually requires expenditure which is how I rationalize it.

From my perspective the threat of me dying from Alzheimer’s is a lot higher than the threat of my dying from climate change phenomena so I fold year round…. but I do it with a bit of a guilty conscience in the summer.


Top
 Profile  
 
 Post subject:
PostPosted: Mon Nov 26, 2007 6:58 am 
Offline

Joined: Fri Aug 10, 2007 6:10 am
Posts: 65
Location: big u.k
i have my system set up to use one of my 2 cores at 100%

this means i can set cystal cpu id to only clock up the cpu when the TOTAL cpu usage gets to 55% thus allowing my amd system to run at just 0.9v and 1ghz most of the time whist still crunching numbers for F@H. this negates any worries i have about running up the electric bill and as for deadlines i have it set up to do large WUs, i only have the pc on about 5-6 hours a day when im using it and never missed a deadline.


Top
 Profile  
 
 Post subject:
PostPosted: Mon Nov 26, 2007 8:38 am 
Offline

Joined: Wed Nov 21, 2007 2:09 pm
Posts: 74
Location: Germany
cvinden wrote:
From my perspective the threat of me dying from Alzheimer’s is a lot higher than the threat of my dying from climate change phenomena so I fold year round…. but I do it with a bit of a guilty conscience in the summer.

Very short sighted. VERY very short sighted. Puts the rest of your post in quite a different perspective.


Top
 Profile  
 
 Post subject:
PostPosted: Mon Nov 26, 2007 8:44 am 
Offline

Joined: Wed Nov 21, 2007 2:09 pm
Posts: 74
Location: Germany
Actually, folding calculations would be done a lot faster if all the people contributing their $$$ of higher electricity bill because of that to a fund for dedicated cluster systems with highly efficient hardware parts instead of let the power suppliers making a fortune out of it. The whole thing becomes absurd, when people even purchase home PCs only for running such a task. And even using availible inefficient home PC hardware is not the way to it. Government should forbid such things as well as using PCs as heatup or prarasitic loads.


Top
 Profile  
 
 Post subject:
PostPosted: Mon Nov 26, 2007 9:31 am 
Offline

Joined: Wed Nov 21, 2007 2:09 pm
Posts: 74
Location: Germany
In that vein, I researched the list of the Top500 supercomputers (and FAH is IMHO a typical supercomuter task) where a system with 128 processors is ranked 72nd. And that amongst systems with usually more than 20 times more processors.

Now let's talk FAH (statistics taken from FAH webpage):
In total around 1,856,800 PCs with around 182,600 CPUs active calculate an impressive number of 174 teraflops.

That's almost equivalent to the second top of the list Blue Gene/P supercomputer which runs around 167 Teraflops. But only uses 65,536 processors! That's 3 times the power just wasted! And were not talking a mere $500 dollar in total for just one system here!

And last but not least, 12 of those ranked 72nd supercomputers would also result in around 189 teraflops (= 12 x 15811 Tflops), with only 12 x 128 = 1536 CPUs. Less than 1% of FAH.

Go on to burn precious energy. Burn it!


Top
 Profile  
 
 Post subject:
PostPosted: Mon Nov 26, 2007 9:48 am 
Offline

Joined: Tue Sep 20, 2005 6:55 am
Posts: 5085
Location: UK
I would probably agree that the cost/benefit ratio for Folding on CPUs is not very good. but nowadays they contribute less than 12.5% of the overall TFlops, and hopefully will eventually be phased out in favour of the PS3. also:

http://folding.stanford.edu/English/FAQ#ntoc8

Quote:
Why not just use a supercomputer?

Modern supercomputers are essentially clusters of hundreds of processors linked by fast networking. The speed of these processors is comparable to (and often slower than) those found in PCs! Thus, if an algorithm (like ours) does not need the fast networking, it will run just as fast on a supercluster as a supercomputer. However, our application needs not the hundreds of processors found in modern supercomputers, but hundreds of thousands of processors. Hence, the calculations performed on Folding@home would not be possible by any other means! Moreover, even if we were given exclusive access to all of the supercomputers in the world, we would still have fewer computing cycles than we do with the Folding@home cluster! This is possible since PC processors are now very fast and there are hundreds of millions of PCs sitting idle in the world.



Top
 Profile  
 
 Post subject:
PostPosted: Mon Nov 26, 2007 10:06 am 
Offline

Joined: Wed Nov 21, 2007 2:09 pm
Posts: 74
Location: Germany
That article is basically saying "we just screw our world and don't give a shit about performance per watt because we can achieve a better peak if we do so". There was already a similar short sighted comment in this thread.

How much energy does a PS3 need when calculating a FAH task? How much is that compared to 0.35 GFlops/Watt of the BlueGene/P? It's around 0.3 GFlops/Watt for the PS3 as I found now... Well, why not just screw the last 10% from the computers then and get a few more PS3 clusters...


Top
 Profile  
 
 Post subject:
PostPosted: Mon Nov 26, 2007 4:10 pm 
Offline
*Lifetime Patron*

Joined: Fri Apr 04, 2003 6:21 pm
Posts: 4247
Location: Undisclosed but sober in US
Mohan wrote:
......
In total around 1,856,800 PCs with around 182,600 CPUs active calculate an impressive number of 174 teraflops.......

That's almost equivalent to the second top of the list Blue Gene/P supercomputer which runs around 167 Teraflops. But only uses 65,536 processors! That's 3 times the power just wasted!......

We know for a fact that all these cpus are of equivalent MIPS and electrical usage?

Is anyone willing to pay to run Blue Gene month after month? Basically insurance works like this, it spreads the pain to everyone, else any "one" probably won't make it.

I'm sure they Stanford would listen to any viable alternatives, as long as they don't involve updating their documentation. 8)

We tried to address this when people started stating their points and wattage per day in another thread.

_________________
People who put money and political ideology ahead of truth and ethics are neither patriots nor human beings.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 44 posts ]  Go to page 1, 2  Next

All times are UTC - 8 hours


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group