FAH Crisis?

[Adak]

It started with the announcement of the changes upcoming in the BA threshold. Now a great deal of frustration and dissatisfaction, is being aimed at both the FAH forum (FF) moderators, Pande Group (PG), and/or Vijay.

[H]ardOCP and EVGA are mainly affected by this change, and they are both VERY dissatisfied with PG. The thread on the FF is now 28 pages long. I'm trying to get a large group to calm down and work on proposed fixes for the problems, but results have not been promising.

Some very powerful folders have started cutting back on the rigs they were folding on, even!

This is a surprise to me. Vijay never got an A+ in donor relations, but I thought he had a C- rating. Apparently, it was more like D+.

This is just for your info. I don't believe we've been impacted with this level of dissatisfaction.

 

[ZL1]

yap .. running mostly dual cores here, i noticed over the last couple months that things have gone down more and more, the more uptime was required to do anything, the more downtime, the bigger the lapse in points .. at this point boinc is becoming a more interesting project, somehow they are more friendly to older, smaller machines

 

[orion456]

Change is inevitable, but it would be better if they gave longer lead times and more info about future plans. People spend real cash on these machines and to be cut off with a few months notice is very discouraging. You need only look at the number of 0's on our front page to see a long list of people with significant points who have stopped folding.

 

[Macaholic]

Have you contacted ChasR and HayesK for their input? I think that would help. You are working against too much emotion right now from other sources.

 

[madhatter256]

What's going on exactly with PG and Vijay?

 

[Macaholic]

What's going on exactly with PG and Vijay?

The blog post found here;

December 17, 2013 by Peter Kasson ·

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!

 

[Adak]

Have you contacted ChasR and HayesK for their input? I think that would help. You are working against too much emotion right now from other sources.

Sent PM's to both.

There is a long history of upset with how PG has handled points, and this has brought it to the boil.

 

[ChasR]

I'm no longer connected to FAH to the degree that I can do anything about this issue. Reading the FF thread is deja vu. Many of the same people with the same positions as the last time BA requirements were changed. As I saw it and still see it, it's Pande Group's project and they know better than I what they need. Donors are free to weigh the cost of their contribution against the relative value PG assigns to it and decide to continue or not.

FAH is still a great project and worthy of our support.

 

[Adak]

This is the compilation of recommendation I made for PG to study and hopefully, act upon. It has been copied to the FAH and [H]ardOCP's forum.

I've added brief notes to put them in context.


1) A clear longer term road map, accurate for at least the next six months, preferably for a year.

8888 Folders will be looking to buy hardware, with folding as one big priority.

8888 Drop the "don't buy hardware for FAH" concept. It will happen frequently. That's just human nature.

8888 Immediate road map for the BA threshold is critical. In addition, include GPU and SMP changes for that same time period.


2) Limit changes to the BA threshold, to one per year. We want to fold BA wu's, but we can't keep up with three threshold changes in less than one year!

8888 a) More info on why the threshold is being raised. Especially "why so much and so often?".



3) A healthy increase in SMP points. The points per day per Watt, are so low the SMP wu's are viewed as not worth the cost to fold them.

8888 a) They aren't worth folding for any GPU folder.

8888 b) They are a huge drop in points, for ex-BA rigs. More than a 60% drop was reported on some wu's.

8888 c) Substantial boost is needed to provide the incentive to fold SMP.


4) Clearer communication, and more of it. There is a broad feeling that PG isn't listening to the donors. The announcements or replies from PG
8888 are typically lacking any specific information. The recent BA threshold increase announcement, is one good example of that.

8888 Typical response: "He wrote a lot without saying much of anything".

8888 Mentioning the threshold increases without giving specifics on the deadline changes. Leaves many BA folders incredibly stressed out.

8888 Several have reported they are no longer folding, just because of the stress they feel about this.


5) We need a non-scientist, but knowledgeable person, to serve as a communicator for the project.

8888 Someone who can speak for PG, on the forum, regularly.

8888 Having the moderators give us guesses regarding PG, is not enough.


6) We need a great deal more empathy from the moderators. Whether their post is concise or verbose,
8888 the attitude of moderator vs. donor, and the condescension we've seen in the past, must stop.

8888 We need to turn the harsh replies, like: "You're not being rational", or "That's a stupid idea", into
8888 "Have you studied that idea in depth?", or "I think there's a better way, what do you think about this ...?".

8888 The moderators give a great deal of info out (A+ in that department), but have fallen into a "donor versus PG or donor vs. moderator",
8888 attitude, far too often. Having fallen into this "us vs them" trap before on another forum, I can attest it's all too easy to slip into it.

8888 Nevertheless, it must be avoided. The moderators are in many cases, the most frequently read "face" of FAH.

8888 We need a lot more positive words of encouragement, (like a smile, but in words), and a lot less of "this is the way it is,
8888 take it or leave it", type of replies.



The above were compiled from both PM's sent to me, and from the BA threshold thread in FF, and on [H]ardOCP.

Your comments are welcome.

 

[JackNSally]

I'm a long time folder. Change has come over the years. Upset about change? So should we all should be folding unirocessor? Go back to getting 1-20ppd? Ppd/watt? So an old 386 should have points adjustment to ppd/watt of smp and/or BA? Change is inevitable. Imbalances may occur in points. But at the end of the day it's the science. Points are an added competitive bonus.
Just my thoughts.

 

[ZL1]

Jack, although I fully agree change is inevitable and more importantly necessary for any evolution to take place, my issue with this years (last years) changes is that dual cores are becoming useless (if not run non stop a 3ghz c2d can now barely finish the WU in time, let alone score well) and thats unfortunate since they can still do some good in my book

 

[JackNSally]

I understand. I do know this. I have c2d still running. The issue mentioned originally was changes in BA, i do wish smp had a bit more to offer dual cores while not penalizing quad+.

 

[dfonda]

I like the implementations... What other reason would I have to upgrade?

Make it 64 cores I'm fine with that. I'm in heaven, I just dropped everyone off my threats list(At least till the next update...) That pesky Adak is always sneaking up....

 

[ihrsetrdr]

I like the implementations... What other reason would I have to upgrade?

Make it 64 cores I'm fine with that. I'm in heaven, I just dropped everyone off my threats list(At least till the next update...) That pesky Adak is always sneaking up....

Dave, I like your style.

Just for perspective: The quest for tethered vesicles, a WU that was worth an astounding 364 points!

I can only guess how much hardware I've gone through, since those days.

 

[Macaholic]

I can only guess how much hardware I've gone through, since those days.

For me their is no need to guess. Just step into the attic and, WHAM! (Tinker units) AMD Dual Socket A servers, (QMD units) Intel Dual Xeon ASUS PC-DL Deluxe servers, Six slot GPU ASUS Wall Street Quartet workstations, etc. You all remember those days, right? The list is staggering. The project moves on.

 

[RoXQi3x]

I like your styles, guys. As a GPU"kiddie" I dare not complain to much on 4p whiners, they have spent a lot compared to me (so far), but I too know some days those GPUs will go out of fashion.
Thanks for the insight on ancient history, and thanks for being veterans with a healthy perspective on development!

 

[Adak]

There has been no response from PG of any change (especially in the communications from PG to the donors).

Many of the BA (bigadv) folders on [H]ardOCP and EVGA are "on strike", choosing to crunch for World Community Grid, work on other BOINC projects, or simply turn their systems off.

I wish they were more patient, but I have been unable to persuade them to wait further. They might return after awhile, because they are so proud of their #1 or #2 ranking in FAH. Impossible to tell, atm.

 

[dfonda]

There has been no response from PG of any change (especially in the communications from PG to the donors).

Many of the BA (bigadv) folders on [H]ardOCP and EVGA are "on strike", choosing to crunch for World Community Grid, work on other BOINC projects, or simply turn their systems off.

I wish they were more patient, but I have been unable to persuade them to wait further. They might return after awhile, because they are so proud of their #1 or #2 ranking in FAH. Impossible to tell, atm.

Thats actually probably the most effective course, to take, to show their displeasure. The drop in folding power will be noticeable to fah and speaks louder than words. If I had a 24 core machine about to be obsoleted, I might choose to do the same. Though I would just switch it to a Rosetta client, so it would still be folding. No one can argue that Rosetta and Boinc points system causes less fuss.

 

[RoXQi3x]

Guys, look at Adaks actions here. Aren't this kind of dedication to people and the cause, as well as patience and tolerance - AND project experience - just The Right Stuff for that new position at PG they mentioned "President for Donor Relations", wasn't it?
I say

ADAK FOR PRESIDENT! "Our man in the fold"

 

[dfonda]

Guys, look at Adaks actions here. Aren't this kind of dedication to people and the cause, as well as patience and tolerance - AND project experience - just The Right Stuff for that new position at PG they mentioned "President for Donor Relations", wasn't it?
I say

ADAK FOR PRESIDENT! "Our man in the fold"

You are absolutely correct! I agree with Adak 90% of the time(There was that one time with the nudity/Booze and Bunny Rabbit ears I objected to...the ears I mean). In this instance though we are just a bunch of hobbiest. I just think running fah is an expensive hobby and I wouldn't want to suggest that I ever know best how to spend others money.

I get disturbed much more about how often other teams members test out their overclocks on BA Wu's(Or any wu's). Talk about wasting science...

Its a forum here...I could just as easily argue for Adak's POV. Given another day and a different mood...

Adak rules!!!