Change in BA requirements

[Adak]

We need calm rational discussion, and patience, not "it's been over a week and I don't have specific answers to this problem!", hysteria.

no, three weeks...no, three months...no, three years...no...

Yeah. not exactly hysteria.
well overdue? YES.

edit PS

@ alias: The "top down" attitude is very efficient,

yeah, it has worked so well in command and control economies whenever and wherever it is tried.

In times of trouble, people need a stronger leadership. In sports, you have the coach or the captain, calling the play - you don't take a poll. In a war, you have the generals and commander in chief. Again, there is no poll taken on how the war should proceed.

You and I can't lead FAH, because although we fold, we have no experience directing a distributed computing project. Not to mention that we are not research scientists at a major university, with our own lab, servers, grad students, and assorted associates around the world.

FAH will be lead by PG, and we can make input on some matters, but it will be lead by PG. It MUST be lead by someone in PG.

[tear]

Because there is no way that we are going to solve several problems, all at once, in just one thread. Let's keep this thread for the BA threshold change problem. Let's enumerate other problems in a thread or threads of their own, and remember the value of patience. These problems were not created over-night, and they won't be solved over-night, either.

Adak, I still remember the value of patience back from 2006 and FCF site. Am I doing good? Can I have my medal yet?

Sarcasm aside, the approach you're proposing is, in principle, absolutely fine.
In healthy organization it's, hands down, the right approach.

What old-timers know, however, it won't lead anywhere around here. It's like a goddamn deja vu.

People get pissed. There's some skirmish between FF administration (to this day I have no idea
who they really represent) and donors. Vijay steps in and says he will take action, then disappears.
Sometimes a band-aid gets slapped, at other times not. Life goes on... until next crisis.

Nothing in this thread tells me this time will be any different.

"But they said they would hire donor relations person!"

Until it happens, I have no reason to treat PG seriously (again, I'm prejudiced given their past record).
Consequently, until said person (if and when hired) is given appropriate prerogatives, we can forget
about any necessary change -- otherwise he'll be just a marionette (with someone else pulling the
strings, if at all).

Have you seen Pulp Fiction? What Vijay should have done is saying something akin to:
"I just want you to know how sorry we are about how ****** up things got... between us and Mr. Wallace."
Three simple words -- "I am sorry". Instead we're witnessing another round of politics and ignorance.


P.S.
This post has been preemptively recorded so evidence of tampering, if any, can be provided.

[Adak]

Adak, I still remember the value of patience back from 2006 and FCF site. Am I doing good? Can I have my medal yet?

Sarcasm aside, the approach you're proposing is, in principle, absolutely fine.
In healthy organization it's, hands down, the right approach.

What old-timers know, however, it won't lead anywhere around here. It's like a goddamn deja vu.

This is a big surprise to me. I'm not that young - Overclockers Australia were the #1 team in FAH when i started folding. I've just stayed out of most FF discussions.

Now we have a choice. We can either try one more time to enumerate our problems, and work them out, or we can do nothing. I'm a fair procrastinator, but not good at "do nothing". By "work them out", I mean work toward a practical solution, which PG can study, and possibly implement. Maybe nothing will come of it, but it doesn't hurt to try.

As a group, and individually, we'll have to let that frustration from the past, go. We need to be calm, and focus on the problem, or the discussion will go straight off the tracks, and be useless to PG or anyone else.

If you want to do that, why don't you start us off: what is problem #1, the most frustrating problem? Let's focus on it, and see what good input we can get from the rest of the readers in this thread, as well.

Pulp Fiction eh? I've seen it. The robbery of the cafe scene ("everything's OK, honey-bunny"), seems most appropriate. Be calm, or else!

[Viper97]

What was 222,988 active machines on Jan. 6, 2014 is now down to 220,854 as of 3:30 AM PST. A loss of 2,134 machines.

Remarkable since on my team the folding for dollars normally bumps up the active machine count. Between all the other DC projects out there and coin mining it won't be long before this reaches the tipping point.

My recommendation is that something be done in announcements and get this firmed up now rather than later. Let the fallout dust what ever it touches. I think folding right now is suffering the 'death of a thousand cuts' due to inaction and or the inability to come to grips with the problem at hand.

It may be too late for a PR person. Especially when the Captain of the ship is busy looking through the spy glass while the stern is sinking. Somebody needs to man the bilge pumps asap.

[troy8d]

Thank you for the suggestions. I need to consult with the rest of the F@H leadership, but I think the likelihood is that we'll want to move sooner rather than later.

I should also point out that one could equally conclude that BA work units are overvalued rather than that SMP are undervalued. A number of people have made this point, and it is not clear (and not entirely in my hands) which direction the points scheme will move. I would personally be in favor of a larger realignment, but that is of course complicated.

Historically, when we first introduced SMP work units, we used a Mac Pro as the benchmark machine. At the time, that machine was relatively powerful, and there was donor concern that a "high-end" machine was not the best choice of benchmark. It is good to remember history, but perhaps we have been over-mindful of it.

I would argue that it is more likely that BA WUs are overvalued rather than SMP WUs undervalued, but it is sacrilegious to suggest that anyone receive less points.

I believe that bigadv has become as popular as it is because of the wide point disparity per watt compared to other types of folding. A reevaluation of the bigadv point premium may be necessary if PG finds the value they bring to the project greater than the 20% initially targeted.

Regarding the optimal benchmark machine, I think your biggest problem is attempting to maintain a single benchmark machine. I know its the most theoretically sound approach, but actual results have proven numerous issues make benchmarking more complicated than that. Keep in mind that the points system will not be subject to peer-review, and ultimately its relevance is determined by the donors.

With the limited information available I can't tell you what is the optimal decision moving forward, but I would just urge you to make sure you consider the long run impact as well. Bigadv folders aren't your typical donor and are, perhaps, more aptly characterized as enthusiasts, fans, or advocates (think Apple users before iPod/iPhone made it more mainstream). Though the nature of some of the posts here may be off-putting, its indicative of their passion for folding and the open dialogue is a good thing. Its when they become indifferent that it is time to be concerned.

[troy8d]

@troy8d, I noticed the 17% drop off between SMP and BA points. What system produced those? Bill1024 had far different data:

I have never been a bigadv folder (and have no intention of becoming one). I was using numbers provided to me from a trusted source. The results I used were from an AMD 4P 6180 (2.4GHz), but I'm working on gather additional data from other users and other machines.

[mdk777]

Though the nature of some of the posts here may be off-putting,

As a group, and individually, we'll have to let that frustration from the past, go. We need to be calm, and focus on the problem, or the discussion will go straight off the tracks, and be useless to PG or anyone else.

Can PG run this program without ruffling folders feathers and without making some mistakes? Of course not. They can use our feedback, to highlight any corrections that are needed, but they can't use either our "moderator tampering" paranoia, or our screaming like a bunch of 10 year old girls.

We need calm rational discussion, and patience, not "it's been over a week and I don't have specific answers to this problem!", hysteria.

For heaven's sake! This is not Armageddon, and the problem will be handled. Will you or I get everything we want? Probably not. When was the last time you got everything you wanted, and it also was exactly what everybody else needed and wanted, too?

OK,
Here is the problem with these posts that imply a lack of patience and calm reflection are a problem.
This post was submitted 3 weeks ago.

Dr. Kasson,

Thank you for the announcement today. I really appreciate hearing from your group from time to time, even if there is nothing new to report.

With that being said, I cannot help but be saddened by this recent announcement. Due to the ability of many server grade computers being able to achieve the current deadlines, I am wondering why there will be a change in the requirements for -bigadv folding.

Are they deadlines for the current projects going to be tightened?
Are new, larger project going to be released?

I remember the core requirement increase 2011 when the core count restriction was changed and the outcry that followed. I understand the reason for that change was to eliminate the single chip -bigadv users as the client was never designed for them. ( I myself had a computer that was built for this sole purpose.) This time, however, I don't see the rationale behind the core increase.

What I'm trying to get at is I feel that I (and others) need more information about the roadmap of the project so that I (we) can plan our future donations.

Thank you for your time,

Erik (bobc36 of folding@evga)

Now, 3 weeks later, there has not been a single reply to the question of why?

I am wondering why there will be a change in the requirements for -bigadv folding.

Now many have attempted to speculate, and guess for PG.
But there is still no answer from PG.
I myself made some speculation for the timing of the two announced and a third anticipated (by Dr.Panda)
Edit: Even 7im has said that he does not understand the need for a two step change much less a three step in a year approach.

So remaining calm and focused on the problem at hand:
re: the change in BA requirement announcement

When do you think is an appropriate time to expect a response to these questions.
We are already approaching the mid-way point to the first date.

The appearance of running out the clock is pretty obvious.
It has nothing to do with demands or hysteria. It has everything to do with how a rational person interprets the appearance.

[7im]

...snip

Now, 3 weeks later, there has not been a single reply to the question of why?

No need to reply when the answer was provided in the original announcement.

We have a policy of periodically re-evaluating the bigadv program, including the threshold required to run bigadv projects. It is the intent of bigadv to match large and resource-intensive work units with some of the most powerful machines used by FAH donors. This "most powerful" line naturally advances with computing power. To date, bigadv has been a CPU-based program, and with increasing numbers of CPU cores and power of those cores, we have decided to lay out a roadmap of bigadv threshold changes for the next several months.

Feb 17 (two months from today): bigadv threshold will become 24 cores
Apr 17 (four months from today): bigadv threshold will become 32 cores

We want to give advance notice of these changes, and we recognize that change is not always welcome or comfortable. We should also emphasize that the science performed by donor machines is valuable in all parts of the FAH project, and part of the change in bigadv threshold is because we would like to encourage moderately powerful machines to help boost the capabilities of non-bigadv SMP projects where we do a lot of this science.

We also recognize that core count is not the most robust metric of machine capability, but given our current infrastructure it is the most straightforward surrogate to evaluate.

Thank you once again for your generous participation in the Folding@Home project!

I read this as kasson saying that due to the increasing cpu speeds and core counts (and the number of those high end machines built), kasson doesn't need as much power to do his research, and other parts of FAH (SMP) does need that help.

This was pointed out back on the first few pages of this thread 3 weeks ago.

As for implementation, a 2 step plan was to make up for short notice, which is lame, IMO on both counts. Doing a two step plan actually works against that idea, because the first step comes 2 months sooner. If they had just done a 4 month warning and 1 step, it would have been less salt on an already sore topic.

I've also seen a 17% drop in points number discussed. Sounds about right, since BA was, by design, given a 20% premium over SMP for the additional resources BA used.
SMP FAQ: What are bigadv Work Units?
http://folding.stanford.edu/home/change ... ork-units/

What else would you like to know?

[Bill1024]

[mdk777]

What else would you like to know?

You have merely repeated your best guess.
Bruce has his.
I have mine.

What donors asked was for PG to explain THEIR rationale.

I could explain why your interpretation makes more or less sense, but that is entirely beside the point.
As you correctly note, for three weeks donors have continued to debate a wide range of issues around the thinking and motivation.
I happen to think many of those questions are valid. Hence, as I think well above a majority of donors have asked, it would be helpful to have more information directly from PG.

[7im]

What else would you like to know?

You have merely repeated your best guess.
Bruce has his.
I have mine.

What donors asked was for PG to explain THEIR rationale.

I could explain why your interpretation makes more or less sense, but that is entirely beside the point.
As you correctly note, for three weeks donors have continued to debate a wide range of issues around the thinking and motivation.
I happen to think many of those questions are valid. Hence, as I think well above a majority of donors have asked, it would be helpful to have more information directly from PG.

PG has explained as much as they are going to explain for the time being. If the why and how are not self evident, I cannot help. Also, as kasson noted, more details to follow.

And while some of the scrooges around here begrudged the lack of response over the Holidays while most Universities all but shut down, I do not. Nor was there a lack of response. You just didn't get the answers you wanted.

I would have liked a response to my request for ALL BA level hardware to get BA level points, Windows, OSX and Linux, not just Linux. Equal pay for equal work, right? Granted, SMP work units may not be as big, use as much memory or bandwidth, but if I have a 64 core Windows server that I would like to fold on, I would like to get some kind of premium for that, even if just half the BA premium. Seem like that would help churn the backlog of SMP work more than upsetting the low end BA people with a core hike. Waiting patiently...

[Adak]

@ mdk777: It sounds like you want a new points system, or an adjustment to the current points system. You also want to know why the change was made in the threshold.

The points change I believe we should work on, and present it to PG for consideration.

The reason for the threshold change was roughly explained - they want only the most powerful systems to fold BA, and that is a moving standard.

I'm sure it won't have a lot of us dancing on the ceiling, but let's work out a points adjustment so ex-BA folding rigs, can still make decent points, folding SMP wu's.

If GPU's can outfold cpu's, then we need to consider increasing further their points. PG wants, and we want, more points for more work in less time, times a scientific factor (sciK I call it).

It's an idea that everyone understands, and also encourages donors to upgrade to the best folding systems.

[mdk777]

@ mdk777: It sounds like you want a new points system, or an adjustment to the current points system. You also want to know why the change was made in the threshold.

Nope.
It really makes no difference to me whatsoever.

If PG wants 3 separate(smp,GPU,BA) or one point system doesn't matter.
If PG wants BA to represent the top 10%, 5%, or 1% doesn't matter.

What matters is that they should be able to explain and defend how they come to the decision.
I don't care at all about the outcome.
I really believe the PROCESS is critical to donor morale,I really believe the PROCESS is critical to the credibility of the project.

I know I am being redundant. But you and 7im keep trying to move off this central point.

Nowhere did I see an announcement that PG would continually adjust the core count to insure that BA machines would represent some arbitrary and capriciously predetermined % of smp machines.

We have a policy of periodically re-evaluating the bigadv program, including the threshold required to run bigadv projects.

Is not a defined rubric. It is not defined metric that anybody can use.

If this always was the criteria, I would not have any problem with constant adjustment.

"Hey, we are going to reward the top 1% of the fastest machines with a 20% bonus. Since this is a constantly changing, you will only know that you are that top 1% when you receive the bonus. It will be determined in real time by the average return time over the preceding 24 hours. Good luck competing and speeding up our science at the same time."

I am totally fine with that.
Nope, I don't have any problem with any combination or permutation of criteria for points discussed by donors in this thread.

What I have a problem with is Bi-annual announcements of change without any underlying logic, predictability, or accountability.

Saying that donors should, or could, or would have been able to make a guess about the announcements underlying logic and rationale just doesn't cut it for me.

Why was the donor advisory Board a waste of time?
In my opinion, because it added a layer of non-communication.
Instead of opening the process, it actually added a second tier of clandestine discussion. (which I guess just wasn't all that productive anyway)

Why are donors angry? because of points?
NO

they are angry because they were blindsided. They are angry because in good faith they took thousands of their limited resource dollars and planned to contribute,... planned to play the game,... as they understood it. Then the rules were changed.
I again have written this at least a dozen times in the last 7 years.

All donors want is clear, predictable and reliable information to make informed choices.

You can try and spin, and miss-characterize the discussion in any way you want. PG failed these donors yet again in this respect. Donors deserve better.

[Adak]

@ mdk777: It sounds like you want a new points system, or an adjustment to the current points system. You also want to know why the change was made in the threshold.

Nope.
It really makes no difference to me whatsoever.

If PG wants 3 separate(smp,GPU,BA) or one point system doesn't matter.
If PG wants BA to represent the top 10%, 5%, or 1% doesn't matter.

What matters is that they should be able to explain and defend how they come to the decision.
I don't care at all about the outcome.
I really believe the PROCESS is critical to donor morale,I really believe the PROCESS is critical to the credibility of the project.

I know I am being redundant. But you and 7im keep trying to move off this central point.

Do you work with other organizations that involve you in their decision making process? Do you know the decision making process for any other BOINC project?

Because I have crunched for several different BOINC projects, and NONE (ZERO) of them have ever involved their donors, in their decision-making process. That experience goes all the way back to crunching for United Devices, before FAH was created.

So I believe you are asking for something you know you can't get. Why are you doing that?

Nowhere did I see an announcement that PG would continually adjust the core count to insure that BA machines would represent some arbitrary and capriciously predetermined % of smp machines. If this always was the criteria, I would not have any problem with constant adjustment.

It WAS part of their initial announcement, and they have repeated that warning, on each change to the BA threshold.

What I have a problem with is Bi-annual announcements of change without any underlying logic, predictability, or accountability.

PG is "calling the plays", here. Not us. We are executing the plays, but the coaches are the ones calling them in to us.

Saying that donors should, or could, or would have been able to make a guess about the announcements underlying logic and rationale just doesn't cut it for me.

It is arbitrary, It could have been handled better -- I'll agree, but it's arbitrary from our "coaches", and it's the "coaches" who are the only ones able to see the entire field of play here, both at ground level, and more importantly, from way up high in the coaches viewing office.

This is a cellular protein research project. It needs to be lead by protein research scientists. Doesn't that make sense?

[mdk777]

This is a cellular protein research project. It needs to be lead by protein research scientists. Doesn't that make sense?

No, apparently they have neither the time, nor the inclination to administer. The administration of the point system, the administration of donor relations has indeed been the weakest link. From what I gather, they excel at cellular protein research.

So,no it makes no sense whatsoever to continue to fail in the same way, over and over again.

per your analogy:Do you know any NFL team owner who also coaches? I don't .