Why did you stop folding (F@H) on your C2D/C2Q/P4D?

Status
Not open for further replies.
A

aldamon

Supreme [H]ardness
Joined
May 24, 2000
Messages
6,672
We are having an interesting discussion over at the Distributed Computing forum and I thought I'd post this thread in the Intel folks. We are looking at the participation numbers in the [H]orde and of the 11,475 people or IDs that have contributed in the past, only 1,151 are currently active. So if you have folded in the past, why did you stop? I'd think with the massive folding power the C2D and C2Q have, the numbers would be increasing. Any thoughts?
 
People don't like the project, people think it's a waste of time/money/electricity, people don't see a direct benefit to themselves, people don't care, people don't see any tangible results to them, people don't care, andddd people think it's a waste of electricity with no "real" results.

;)

Oh, and I fold, heh :p with my lone e6400 :(
 
I understand why those issues would be important to someone who has never folded, but I'm talking about folks who have already contributed. Those issues didn't stop them before so I'm curious why they aren't folding now with hardware that is faster :)

gwai lo said:
Oh, and I fold, heh :p with my lone e6400 :(
Click to expand...

Why are you frowning? That's an excellent folder.



 
I'm :( about the "lone" part :p

Oh, and I think those issues apply to people who have folded. I imagine after a certain point, they just don't see any benefit from running it...most people like to see tangible results instantly. Kind of like cheap PSUs and quality PSUs...people keep buying them because the guy who bought a coolmax yesterday is doing just fine..they don't care that PSU issues are long term and the guy might not be doing so great in a couple years.

..but then again we don't really know if anything's going to happen in a couple years. ><
 
I'm one of those people who thinks there are no real results... I mean how many times does the same protein need to be folded before 'results' are found? Maybe I'm wrong though because it has been months since I ran F@H on my computer.
 
If it is that easy to find a cure to aids, why do they need our computers to do these projects? think about it.
 
bail_w said:
If it is that easy to find a cure to aids, why do they need our computers to do these projects? think about it.
Click to expand...
hah, now that's a bad answer.

There's more processing power if they open it out to the public, think about how many computers are in the world. ;) They do have servers running the project, but a server != god-like-computer-that-has-processing-power-equal-to-the-sun. Actually...any google on a distributed computing project will get you hits very quickly explaining it, in addition to giving specific numbers about processing power.
 
I folded for [H] for a year. I had a little spat of activity recently, but I stopped folding regularly. My reasons were pretty simple:

1) Don't like running computer 24/7.
2) Don't have central A/C and my office gets really hot with the computer on 24/7 at 100% load. (I live in Arizona where it's over 100 most of the spring, summer, and fall.)
3) Running F@H in the background didn't do any good when I was gaming because F@H never got any processor time, and I prefer to turn off my computer when not using it because of reason 1 and 2.

Pretty simple really.
 
I stopped folding because:

#1) I don't like running my computer 24/7.
#2) I noticed a significant increase in my electric bill.
#3) I honestly just stopped caring to do it. I don't know why I should be bothered to pay more than I should anymore. I can't justify it in my head anymore. I understand the cause and I understand the effort but it just doesn't seem like something I should personally continue. I have a few folders left around the house that are folding for another team but still F@H. They will be put out of commission soon.
 
1) Still just a "theory"
2) Nothing has been introduced as a result of folding (read: no drugs or cures)
3) Cost of power and cooling (significant resources)
4) Cost of hardware
5) They have more data than they can process, resulting in pointless backlog of work which may never be processed
6) Points system (cool concept, but has a major flaw in that new folks are years behind in catching up, even if they have more processing power)... it should have either used a ramping algorithm with some weights to the scores or depended soley on month-to-month production... so new folks have a chance.
 
Riddlinkidstoner said:
I stopped folding because:
#2) I noticed a significant increase in my electric bill.
Click to expand...

how much more significant increased in your bill? i have 2 quads and a core 2 duo, and it only increase ~30 dollars a month for running 24/7.
 
bail_w said:
how much more significant increased in your bill? i have 2 quads and a core 2 duo, and it only increase ~30 dollars a month for running 24/7.
Click to expand...

Because I'm using a few laptops and a older P4, I saw an increase of about $45.

Either way, $30 increase is still a lot of money.
 
Old technology use lot more electricity, anyway, i can afford to give 30 dollars a month. besides, it is a good stress program to test my overclocked system.
 
Astral Abyss said:
1) Don't like running computer 24/7.
Click to expand...

When did everyone get the idea that a computer needs to run 24/7 to fold? Is this a biproduct of the points/competition system? The program is designed to use unused CPU cycles and run in the background. The SMP and GPU clients are hogs and can cause problems, but the standard client runs in the background beautifully. My wif'e's computer is a Venice 3200+ and it folds 8 hours per day for her work day and contributes.

bail_w said:
besides, it is a good stress program to test my overclocked system.
Click to expand...

Hmm, I am wondering if maybe some of the systems around here used for screen shots can't handle the task :)

 
eh...don't stress with F@H. Use Orthos' Gromacs core test, not the real thing..
 
bail_w said:
what you mean dont stress with F@H? Not the real thing? what is the real thing? care to explain?
Click to expand...

You can f up the results if your sending back bad WUs. That's the one concern, make sure your PC is stable before folding with it.

Personally I still fold, figure it'll help out in the long run and it's a great way to build comradery with other people on the team and in the community. It's not that hard to put up numbers, I've been folding for around 14 months now and am in the top 400 overall, you can climb the ranks pretty quickly.
 
aldamon said:
When did everyone get the idea that a computer needs to run 24/7 to fold? Is this a biproduct of the points/competition system? The program is designed to use unused CPU cycles and run in the background. The SMP and GPU clients are hogs and can cause problems, but the standard client runs in the background beautifully. My wif'e's computer is a Venice 3200+ and it folds 8 hours per day for her work day and contributes.



Hmm, I am wondering if maybe some of the systems around here used for screen shots can't handle the task :)

Click to expand...

more likely because of the deadlines on WU now, also i know the scores are good and can add some competition to it, but this isnt about who has the most points, it not an ego contest, or was not meant to be but obviously people think it is and they only do it so they can brag about their score.

it would be nice to see some more results from F@H, more drugs and such...
 
ND40oz said:
You can f up the results if your sending back bad WUs. That's the one concern, make sure your PC is stable before folding with it.

Personally I still fold, figure it'll help out in the long run and it's a great way to build comradery with other people on the team and in the community. It's not that hard to put up numbers, I've been folding for around 14 months now and am in the top 400 overall, you can climb the ranks pretty quickly.
Click to expand...

Orthos is not guarantee your system to be 100% stable. In my case, Orthos 8 hours + stable failed OCCT 30 mins test. So if you use F@H to fold and found out that your WUs are invaild = your system is unstable. anyway, Fold or not its up to you.
 
bail_w said:
Orthos is not guarantee your system to be 100% stable. In my case, Orthos 8 hours + stable failed OCCT 30 mins test. So if you use F@H to fold and found out that your WUs are invaild = your system is unstable. anyway, Fold or not its up to you.
Click to expand...

I'm not going to get into a discussion about how to test the stability of a system, since this isn't what this thread is about. Bottome line, don't use folding at home to do it, there are plenty of alternatives. Stanford specifically asks that you fold on a stable system and not to test the stability of your system with it.
 
bail_w said:
If it is that easy to find a cure to aids, why do they need our computers to do these projects? think about it.
Click to expand...

Right, but when I see the same protein being folded over and over, I can't help but feel it's a waste of electricity. Especially when thousands of other computers are constantly folding that same protein.

I'm not saying it shouldn't be hard to find the cure for cancer - but I'm not going to dedicate a machine to process results that have already been processed thousands of times.

F@H to me honestly seems like a giant test (and successful) for possible future distributed projects.
 
F@H is a basic research project dealing with the fundamentals of biochemistry and initiation of fatal diseases. There will be no drugs coming out of this project, but valuable knowledge on molecular behavior that may spawn applied medical research on cures/drugs for the diseases mentioned in a 5-10-20-year perspective. And sometimes basic research turns out to be useful in areas beyond the original project objectives.
 
1) Significant increase in room temperature (AC isn't going to kick on because only this small room heated up).
2) Significant increase in electricity bill.
 
And btw, it's not very likely that Stanford is running the same WUs on thousands of machines. If they were running out of proteins to model they could easily modify some of the existing WUs to model folding in aqueous environments, i.e. one step closer to reality. They've already done that with a few WUs.
 
I know we're geeks and all, but might it be more efficient to save on the power consumption and just cut an old fashioned check to some worthwhile cause?

/runs before the rocks come flying...
 
ND40oz said:
You can f up the results if your sending back bad WUs.
Click to expand...

Yes, you can but the results are verified and if you return corrupted results you get no points. So it's best to have a stable system, not because you're going to screw the project, but because you're doing work for nothing:

__________________________________________________________

http://forum.folding-community.org/viewtopic.php?t=19062

7im
Site Moderator

Obviously, the faster you can turn in the completed work, the better for you (points) and better for the project (science). The data is error checked, and if you overclock too much, the data is corrupted, and you don't get any points. The system tends to be self-correcting, so no need for an official statement. ;)

As a relative benchmark, if your system can pass the Orthos or StressCPU test tools found in the 3rd Party apps section, then your system is likely stable enough for the CPU client. For the SMP clients, back it off 100MHz or so additional.

_________________________________________________________

Volkum said:
1) Significant increase in room temperature (AC isn't going to kick on because only this small room heated up).
2) Significant increase in electricity bill.
Click to expand...

If you just fold when your computer is on anyway, I don't see how you'd notice either. :confused:

SonOfMore said:
I know we're geeks and all, but might it be more efficient to save on the power consumption and just cut an old fashioned check to some worthwhile cause?

/runs before the rocks come flying...
Click to expand...

No rocks, but I'll just say your comment was not constructive either. If you folded with that Q6600 just when your computer was on anyway, you'd be able to contribute a ton of work units.

 
aldamon said:
Yes, you can but the results are verified and if you return corrupted results you get no points. So it's best to have a stable system, not because you're going to screw the project, but because you're doing work for nothing:
Click to expand...

Exactly, you're screwing up the results because they don't verify and then they have to send that WU out again, causing that piece of the puzzle to take longer. It would be better not to do any work at all then turn in a corrupt WU. Someone else could have folded the protein correctly and since other WUs depend on ones that are turned in, that project could be delayed because of bad results.
 
aldamon said:
When did everyone get the idea that a computer needs to run 24/7 to fold? Is this a biproduct of the points/competition system? The program is designed to use unused CPU cycles and run in the background. The SMP and GPU clients are hogs and can cause problems, but the standard client runs in the background beautifully. My wif'e's computer is a Venice 3200+ and it folds 8 hours per day for her work day and contributes.
Click to expand...

Please read my entire post. When I'm not gaming my computer is off. Therefore in order for F@H to get any process time at all I would have to leave the computer on when I'm not using it.

Now, that was on my single core CPU. I experimented with folding on my quad core and it obviously can utilize the extra cores my game isn't using, but I just don't think I'm going to get back into it. I generated almost 90,000 points for F@H in the year I folded, and I'm pretty happy with that.
 
Astral Abyss said:
Please read my entire post. When I'm not gaming my computer is off. Therefore in order for F@H to get any process time at all I would have to leave the computer on when I'm not using it.
Click to expand...

It was not my intention to offend you. That reason was given another time after your post. I just quoted you to try to address the point. Sorry. I failed for your situation.

Astral Abyss said:
I experimented with folding on my quad core and it obviously can utilize the extra cores my game isn't using, but I just don't think I'm going to get back into it.
Click to expand...

That would be great. You might be pleasantly surprised. The Q6600 is a beast folder, even with multiple instances of the standard client. Why not if the cores aren't doing anything anyway?

ND40oz said:
Exactly, you're screwing up the results because they don't verify and then they have to send that WU out again, causing that piece of the puzzle to take longer. It would be better not to do any work at all then turn in a corrupt WU. Someone else could have folded the protein correctly and since other WUs depend on ones that are turned in, that project could be delayed because of bad results.
Click to expand...

I think the percentage of computers that are so borderline stable that they can somehow complete a work unit over the course of 24+ hours or days but submit bad results would be very small. There are too many errors that could happen before the results are sent. This would be for another discussion though. In context of this thread, I would say people who overclock should not be afraid of folding.


 
aldamon said:
If you just fold when your computer is on anyway, I don't see how you'd notice either. :confused:
Click to expand...
There's a big difference between "on" and "100% cpu load" in terms of power consumption, heat, and noise. Cost of living has been rising rapidly in my area and unfortunately quitting distributed computing projects results in a significant savings on my power bill. Not only in directly reducing computer electricity usage but less waste heat cuts down a bit on fridge, freezer, air conditioning etc. When I get a new more efficient fridge in a few months the pc will be my single biggest electricity expense by a wide margin even without F@H.
 
rls669 said:
There's a big difference between "on" and "100% cpu load" in terms of power consumption, heat, and noise. Cost of living has been rising rapidly in my area and unfortunately quitting distributed computing projects results in a significant savings on my power bill. Not only in directly reducing computer electricity usage but less waste heat cuts down a bit on fridge, freezer, air conditioning etc. When I get a new more efficient fridge in a few months the pc will be my single biggest electricity expense by a wide margin even without F@H.
Click to expand...

I concede the point but I hope you're underclocking your 8800 GTS as much as possible when your computer is in 2D mode. It's a power hog.

EDIT: Total system numbers shown. My bad.

 
I still fold a bit in the summer, but only ticking over on a couple of slower servers when the weather gets cooler my home is partially heated by folding.
and this year there are going to be a few more cores in the mix!
 
Electricity, plain and simple. To the guy that said he can afford $30 a month, not everyone has that money to spend on something they get nothing for. Its a good idea and cause, but nothing has become of it and most likely never will.
 
wee96 said:
Electricity, plain and simple. To the guy that said he can afford $30 a month, not everyone has that money to spend on something they get nothing for. Its a good idea and cause, but nothing has become of it and most likely never will.
Click to expand...

Incredibly stupid and ignorant comment.

People generally donate to science and charity without expecting something back, and contributing directly to research through folding is a donation just as good as cash. And the potential long-term reward to the donators and their descendants in the form of cures and proactive measures against fatal diseases is huge. The results of the F@H project so far is reported in the 50 publications described on their home page:

http://folding.stanford.edu/papers.html
 
Ockie said:
1)
3) Cost of power and cooling (significant resources)
Click to expand...

^ This. Maybe when Fall / Winter rolls around I'll consider it again.

I'm sure plenty others feel the same way, its been a hot summer. I bet the number go up towards the end of the year.
There's a good question for you guys... Christmas Spirit + Cold Weather = More folding ??


~edit~ Should have read the whole thread, people have mentioned this already.
 
Status
Not open for further replies.
Back
Top