To all folders "Folding Forum" needs your help

Bill1024

2[H]4U
Joined
Apr 26, 2013
Messages
2,521
Ok guys, you asked for communication. We got it.

PG has a backlog of SMP work units and they need them folded.

Bruce asked for help and he said 'Please aim your bigadv rigs at regular SMP"

Last page of this thread.

https://foldingforum.org/viewtopic.php?f=16&t=25411&p=254213#p254213

Here is the quote
Re: Change in BA requirements

Postby bruce » 41 minutes ago

OK.

Please aim your resources at SMP rather than BA.

Then in really small letters. He wrote

(I've got to admit I feel pretty stupid doing this.)

If we show we can and will respond to a call for help maybe they will reward us with no change in the BA requirements.




I know there will be a drop in points doing SMP.

I am willing to put all my systems 48, 24,24,6,6,4 doing SMP 24/7 to show we are not all about the points.

SMP needs to be folded and we will step up to the challenge if we are asked to do so.
 
Last edited by a moderator:
Im sorry but until they make SMP equal footing (In the PPW catagory) to even GPU's, there is no way in hell my systems will touch them. PG unleashed a beast with the first bigadv, and as such has just let it grow with the introduction of the QRB on GPUs. The SMP's are larger than they were before, and the factors have remained the same, my 3770k will be lucky to get 15k today whereas back 2 years ago a 920 would pull 30k+ on regular SMP, to me this is bullshit and unacceptable


edit: To add, why would I turn my machines to something with a very large loss to my points return, yes its for the science, but your look at (on some systems) over a 50% loss in production for the same power consumption, does this mean -SMP is only worth half of what a -bigadv is research wise?
 
PG set the points up for a reason and they need to adjust SMP if you ask me. He is pretty stupid for doing that IMO.
 
10000000000000 points and a dollar fifty will buy you a cup of coffee.
Maybe if we show we care they will make an adjustment.
 
10000000000000 points and a dollar fifty will buy you a cup of coffee.
Maybe if we show we care they will make an adjustment.

Fat chance.

SMP is a waste of power for 1p setups. The PPD is just pathetic, and just about everybody knows it. They have a backlog of SMP units because of how poor they made the PPD returns.
 
I wants treefiddy to fold SMP.

Hahaha I was going to say a dollafiddy.

Merry Christmas H and a happy new year.

They are now communicating and asked for help.
He even said "please"
Maybe after the WGC Christmas challenge fold SMP for a few days to show we will work with them if they work with us.

Once the lins of communication are open and it is a two way street, they may listen to the concerns about the loww PPD for SMP.

Hey is is a start.

Merry Christmas
Thanks for at least reading my post.
 
No disrespect to bruce but bruce asking people to please switch their BA over to SMP is not the same as Dr. Pande asking the same.
 
Hahaha I was going to say a dollafiddy.

Merry Christmas H and a happy new year.

They are now communicating and asked for help.
He even said "please"
Maybe after the WGC Christmas challenge fold SMP for a few days to show we will work with them if they work with us.

Once the lins of communication are open and it is a two way street, they may listen to the concerns about the loww PPD for SMP.

Hey is is a start.

Merry Christmas
Thanks for at least reading my post.

It ain't our fault the lines of communication are not a "two way street". They've chosen not to listen to our reasoned critiques of SMP's PPD, with lots of math included to back them up, for a long time....and no, they won't magically listen to us suddenly if all of [H] switched over to SMP.
 
No disrespect to bruce but bruce asking people to please switch their BA over to SMP is not the same as Dr. Pande asking the same.

Bruce can suck my sexual organ. Him and 7im can both suck my sexual organ,

feather, delete this if you see fit.......
 
Hey, at least if we switch for a couple of weeks we can prove the points problem to them once and for all. They cannot dispute the data that a team like ours can provide.

Besides, it would be a challenge to try and chew through the backlog and last time I checked we were out of challenges, having just defeated default and doing 2 billion in a month.

How about 1 billion in a month on SMP?
 
If there is a groundswell of support for this, I'll play along.
No point in doing the lone ranger thing though.

I'll watch this thread and see what happens.
 
No disrespect to bruce but bruce asking people to please switch their BA over to SMP is not the same as Dr. Pande asking the same.
Very much so. He's not in position to make such requests.
 
That is not the issue, most /all of us do, and have machines that can very easily do SMP better than most, its the principle behind -SMP having horrid points (especially considering the wattage it pulls in comparison to -bigadv and GPU)

:( you guys dont have any machines that can fold SMP?
 
Hey, at least if we switch for a couple of weeks we can prove the points problem to them once and for all. They cannot dispute the data that a team like ours can provide.

Besides, it would be a challenge to try and chew through the backlog and last time I checked we were out of challenges, having just defeated default and doing 2 billion in a month.

How about 1 billion in a month on SMP?

LOL...given the awful PPD, before that would happen, EVGA would be back on their way to catching us.

:( you guys dont have any machines that can fold SMP?

I have 1p machines that *could* fold smp. Problem is, the points are a joke for the power bill I pay every month. Further ALL the SMP WUs are over 24-36 hours long for an i7-2600 at stock folding 100% and doing nothing else, so I lose use of a machine...and for what? 8-10K PPD absolute tops?

They need BigAdv clients to do SMP because Stanford has gone and made the SMP units ludicrously big and long (larger in TPF than bigadv)....and surprise surprise....the SMP clients online cannot finish them fast enough, and when they do the points are a joke. Only the 2p and 4p and higher server blades normally doing bigadv have enough computational power to calculate SMP WUs in a reasonable amount of time.

Stanford went and made the mess....and unless they do something about the problem of smp's TPF increasing and returns decreasing the problem will not go away. There will just be a monthly plea for donors to take a 75% PPD cut "and please get rid of the SMP backlog" that is too big for 1p SMP clients.

Because last I knew the point of smp was for 1p donors primarily in purpose...which it really isn't in practice anymore. The WUs are too big and hard for 1p to be worthwhile, which is the source of the problem.
 
Bruce is nobody. PG remains silent on the subject. PG has the ability to make SMP worth running. The fact that they don't tells me that they aren't that concerned about it.

The title of this thread is incorrect, and should be changed. Bruce is making this suggestion, and I am with Dookey on what Bruce can do.
 
Many from EVGA are doing this, I doubt you have to worry much about them catching up to you.
A good handful stopped folding and went to crunching.

I kind of agree about Bruce and totally agree about 7im.
Bruce is not the head of the program. Would have been nice to see VJP ask.
Some one know him and want to ask him what he thinks of this and see what he has to say?
I am glad to see a debate and Ideas on how it can be made better.
I think they may listen.
Have to admit we got their attention. What ever you folks decide.
Thank you [H]
 
I kind of agree about Bruce and totally agree about 7im.
Bruce is not the head of the program. Would have been nice to see VJP ask.
Some one know him and want to ask him what he thinks of this and see what he has to say?

I would ask but it would go unanswered like my PM to Kasson and to you over to the FF.
 
I would argue that BA and Big GPU are overvalued, not that SMP are undervalued.
But then, I come from the days where work units were worth considerably less.
 
Regardless, it will be interesting to see the outcome if FAH project really needs SMP work to be done faster. I have no problems either way, I have been doing SMP for quite a while now on 2 machines. PPD doesn't matter to me anymore. The work still needs to be done.
 
Regarding raising core counts, they (PG) just turns Folding@HOME into folding@WORK ... because not many (not all) donors could meet that requirement at home, and kinda "FORCES" addicted folders to either upgrade their rigs or fold SMP. Either way, it is their WIN-WIN situation. However, it could be back fired at them.
Would they care much what donors think? I guess not. If they do, they would already kindly ask folders to switch to SMP. I would be happy to switch all my 48 cores dedicated folding rigs to do just SMP.


To Mekkanic,
Just put your ass in others' situation where they run multiple multi-processors rigs at HOME and pay their freaking high utility bill, then you will understand their frustration.
 
I hadn't read anything on foldingforum since 2008 until I clicked on the thread in the OP, and now I remember why I always avoided that place despite being a frequent donor to the F@H project. The posts from PG and their backers have always reeked of typical lazy scientists (or worse) not understanding reality and the thread in the OP is just another perfect example.

The bottom line is that PG uses points as a proxy for how much scientific research donors are contributing to the project. That's how they provide not only an incentive for donors to fold in the first place, but also an incentive to fold as much as they can, in a certain manner, and for certain projects. This is a problem they created - a systemic problem caused by using points and compounded by substantial WU point inflation over time not to mention schemes like ERB.

PG, like other people on FF and even here, keep saying things like 'the points don't matter because it's about the science' or "work still needs to be done". Well, duh, of course it's about the science and the work. That's why we're contributing in the first place, because it's for a good cause. But at the end of each day, all that donors can actually see is how much PG tells us our work is valued (points) and how much it costs us to build our machines and keep them running ($). Why is it any surprise then that donors care about things like points/$? That's PG's doing, not ours. It is thus not only hypocritical but actually dishonest for PG to then turn around and tell us that we should ignore the incentive scheme they created.

The problem with oversupply of SMP units is simple: PG has provided an extremely strong incentive for donors to move to BA/ERB-focused folding and donors have responded by building multiprocessor machines and shutting down their other rigs. I would imagine PG should be happy with this response! If PG wants donors to move back to SMP folding, even if temporarily, then it is as simple as providing an incentive to do so through the points system - not by threatening their donors with an arbitrary ban on BA/ERB folding or by insincere begging by PG on foldingforum.

Foolish ideas like an arbitrary core limitation restriction for BA are just poorly conceived plans that are poorly executed to boot, and this is coming from someone whose folding rigs should not be affected by the change. PG should focus on fixing actual problems, not knee-jerk band-aids that only demonstrate the chasm between how PG's administrators think and reality.

Please note I don't fold for the Horde (sorry), so my comments should not be taken as coming from a member of this folding team.
 
You guys could protest by switching to BOINC projects full time... :D Looks like a bad time to consider starting up a few older rigs to get the feet wet at FAH. I'm holding off on upgrades and such until I get a better idea what will be best for me long term vs. my budget issues.
 
It should be noted (and I will be editing the title slightly, OP if you have an issue please PM me about it) to reflect that it is NOT a member of PG requesting this, it is merely a moderator at the foldingforum, and not an official request from PG (As far as I know)
 
It should be noted (and I will be editing the title slightly, OP if you have an issue please PM me about it) to reflect that it is NOT a member of PG requesting this, it is merely a moderator at the foldingforum, and not an official request from PG (As far as I know)

Certainly not an official request from PG; however, foldingforum is the venue that PG has chosen to officially communicate with the folding community, Bruce is foldingforum.org's site administrator (not just moderator), and Bruce's statement was in defense of PG's handling of the SMP backlog.

In any case, I think calling it a "request" is also inaccurate since it seems like Bruce's intention with his "request" was just to belittle Bill1024 (hence why I referred to it as "insincere" above).
 
shrae said:
Why is it any surprise then that donors care about things like points/$? That's PG's doing, not ours. It is thus not only hypocritical but actually dishonest for PG to then turn around and tell us that we should ignore the incentive scheme they created.

The problem with oversupply of SMP units is simple: PG has provided an extremely strong incentive for donors to move to BA/ERB-focused folding and donors have responded by building multiprocessor machines and shutting down their other rigs. I would imagine PG should be happy with this response! If PG wants donors to move back to SMP folding, even if temporarily, then it is as simple as providing an incentive to do so through the points system - not by threatening their donors with an arbitrary ban on BA/ERB folding or by insincere begging by PG on foldingforum.

Foolish ideas like an arbitrary core limitation restriction for BA are just poorly conceived plans that are poorly executed to boot, and this is coming from someone whose folding rigs should not be affected by the change. PG should focus on fixing actual problems, not knee-jerk band-aids that only demonstrate the chasm between how PG's administrators think and reality.
Bartender! Pour this man a drink!

shrae said:
Certainly not an official request from PG; however, foldingforum is the venue that PG has chosen to officially communicate with the folding community, Bruce is foldingforum.org's site administrator (not just moderator), and Bruce's statement was in defense of PG's handling of the SMP backlog.
To me it looks like he's usurping the right to be a PG rep and, given lack of supervision from
the project, he gets away with it 99% of the time.

Many times I've seen statements from bruce that, in normal organization, would've been PR
suicide. That tells me that either noone's ever looking over guy's shoulder or they do but ultimately
don't care enough to do anything about it (or don't recognize the problem whatsoever).
I sure hope it's the former.

What saddens me the most, though, is that getting the project to solve problems is like pulling teeth.
Every time people need to get pissed off -- when the hell did that become a prerequisite?
 
I could turn on my old i7-920's and change my 4P to fold SMP, if PG asked nicely and changes the point scheme to make SMP more in line with GPU and BA wu's.

Until that time, my 4P rig will continue to fold BA.
 
W.Feather I have no problem changing the title as long is it is some what respectful.

Up front I fold for two teams, EVGA and The Folding Wolves
A bit team and a 10 member team. Been folding since 2006 or so.
There is a reason I have very few posts at the FF, as many of you know how the attitude of some people can be. But yet we all fold.

I hope things will get better, the point system has never been perfect.
I would like to see a sliding scale of the K factor (the multiplier) for the smp wus.
The faster you return the WU the higher the multiplier.
As far as I know the multiplier (k factor) is set .75 - 3.xx depending on the WU and type.
Maybe that would = the points some BA vs SMP
And the incentive would to be, to build the fastest, most power system you can afford.

I do understand most of your POV as I too have been around a while.

Thank you guys, your guides have saved me a lot of time and helped many of us be better folders.
Than you.
Bill
 
My apologies, I was under the impression that bruce was just a moderator. Still not in agreeance with their current points scheme...its an incentive scheme, as it stands it shows me that PG could give two shits if SMP is being folded.....their points have depreciated in the last 2 or so years, while GPU and bigadv have increased (Or atleast stayed within nicer point values)


Bill - Changed it early, any issues with the new name please let me know
 
I am a sucker for disasters and will switch my single BA capable machine to SMP for a while, the machine is at work making the power cost 0 to me.
Getting max points out of it is nice, but I do not mind trying to save idiots.
I'll probably stop when mekkanic has a good lead on me and I can mow his behind again.

The current situation is a demonstration of epic lack in communication abilities from PG, and it looks like they have a history of this.

Switching to SMP to remove the backlog will not help anything, SMP is underrated and and if we clear the backlog now, it will just become a problem again in X months because people do no want to fold SMP as it stands.

The point scheme is and will never be perfect, because there are a lot of factors to balance and I think a perfect balance is impossible to achieve.
Unfortunately the current balance makes the donators flee from SMP and PG has to step in and fix this.
 
To Mekkanic,
Just put your ass in others' situation where they run multiple multi-processors rigs at HOME and pay their freaking high utility bill, then you will understand their frustration.

I have been there my friend, back when dual SR-2 were the norm to fold for BigAdv. I reallocated my hardware to where I could use it the most after they changed the standards for BigAdv, and started folding again after a brief hiatus. My two L5640 CPUs are now doing SMP folding.

@runs2far, mow all you want. Not gonna hurt my feels :D
 
Last edited:
Back
Top