Change in BA requirements

[HaloJones]

Your view then is that when BA was setup and it was said that the requirements would change that was all the warning anyone needed? That Stanford can now change the requirements based on how it was initiated? Fair enough, but don't expect too many people to then contribute to that BA program.

Nothing has changed on the multi-socket hardware recently. And Intel and AMD do actually have quite well publicised roadmaps for their server hardware because the big customers demand it. Right now, there is no obvious reason to change the core count requirements once let alone twice in such a short time. All people are asking for is some rationale; is that really too much to ask for? Your defensiveness seems uncalled for in the circumstances.

[Grandpa_01]

Would it not be better to set a date and keep it. A roadmap would be far superior in my opinion to we will be evaluating this again someday when something happens which is pretty much useless to a person who is considering there options for a folding investment.

Something like on on ( July 1, 2014 ) we will be reevaluating the core count and deadlines for bigadv WU at this time the core count / deadlines may change.
On (December 31, 2014 ) we will be reevaluating the core count and deadlines for bigadv WU at this time the core count / deadlines may change.

Now if I was considering a upgrade to bigadv I know exactly how long it will be before it may no longer qualify for bigadv and I can then make my choice as to what I am going to do.

Is that really asking too much.

[patonb]

Screw a roadmap, just setup wu that fill the gap between 2/4 and 32 cores.

I'm sure that it'd be beneficial to research that explores systems larger than smp but not as complex as BS units.

Given server equipment isn't readily overvclocked there's probably tons of hardware idle that could do 12 requirements, even I have it due to last upped core count.

The other thing I not sure happens is why my equipment can't get equal points through out wu types. Technically, shouldn't I get the same ppd given I do smp or BA if both have the same importance?.

[Grandpa_01]

bruce gave an explanation as to why bigadv has a greater science value earlier, although I do not agree with the =pay for =science scenario to me it makes things lopsided and creates points disparity (the proverbial carrot) I do understand the theory behind it and it does work as far as enticing people to do a given type of work, but it does create some pretty big problems in other areas of F@H.

As far as having a different class of smp they already have it if you have a MP rig and run smp you will get the 85xx through the 88xx a3 WU's most if not all of the time which are 166K to 167K atoms and have a very short deadline comparatively to all other smp WU's and take quite a while to run even on MP rigs. I am pretty sure they would not have to do much to these to accommodate what you are suggesting.

[bruce]

Off-topic reply to an off-topic aside:

* I should add that I am not affected by this particular announcement, but I can imagine the frustation of those who are. (The effect of the current dearth of Core17 WU's, on the the other hand, ...... )

viewtopic.php?f=16&t=25438&p=254162#p254162

[tear]

Umm, you're missing the point.

Two months? Try two years! Two years ago it was clearly stated that BA was an experimental program and that it would be reassessed from time to time.

Of course, but that does not constitute advance notice. Would you still say two years notice was given if changes were announced today, effective tomorrow?

You surely must be joking.

When you give advance notice to terminate, say, your employment contract, does it look like:
(a) "One day I'll leave, not sure when, but I'm letting you know now."
(b) "I'm leaving two weeks from now" ?
Think hard.

Notice given on Dec 17, effective Feb 17 are two months worth of notice. You can slice it or dice it any way you want.

And a year ago, it was changed. And again the message about future revisions were to be expected was repeated.

I would call two years worth of notice to be very upfront. Not at all last minute as you seem to think.

What a joker you are, you!

As for a better roadmap, yes, I would like to see that as well. But fah's ability to do that is very limited because we're all a slave to the PC hardware development cycle. Fah can no more predict the exact date the next GTX 790 will hit the street than NV can. Or when NV might get their JIT software working and double the speed of GPU folding. Anyone here know those dates?

Like I explained earlier, I'm not asking about detailed roadmap but for something a donor can rely on.
Specifically, a de-facto guarantee of bigadv-usefulness of donor's hardware for a period of time. 6 months are not too much to ask, are they?

Proposed bi-annual adjustment opportunity (with 6 months notice) would realize said guarantee in professional and coordinated fashion.

[tear]

[
     EDIT: there is also way of telling who's folding bigadv units -- you can use passkey to subsequently retrieve
          donors' e-mails -- why not communicate over e-mail with bigadv donors?

Agree to everything your wrote, except the email thing for BA donors; while it would be good for active BA donors the same information is even more important for "want-to-be BA donors" like myself.

Absolutely, agreed. With passkeys in place, one can communicate with all donors, not just bigadv donors [and I would very much like to see user registration being integral part of FAH].

WCG has been doing it and I genuinely love it. You can clearly tell these guys are not indifferent and spend more resources on overall "donor satisfaction" than FAH.

[7im]

...

6 months are not too much to ask, are they?

No. Neither is this...

1. Anyone building a borderline BA system (16 cores or slightly more) in the last 6 months knowing that core counts get raised from time to time, and that core counts haven't been changed in quite a longer time, was short sighted or given bad advice. I certainly didn't build any GTX 480 GPU farms on the cheap in the last 6 months, did you? And I certainly didn't build any AMD GPU farms knowing that NV JIT would double NV GPU performance at some point in the future, did you?

2. People assumed incorrectly that all BA-16 work units immediately stop getting assigned 2 months from now. While not yet confirmed, things will likely happen that same as the last BA change (BA-8 WUs were available for *many* months after the last change). You got 2 months notice, plus however long the BA-16 WUs last, plus a 2 year heads up, plus a strong reminder a year ago.

Slice that any way you want, but that is not a lack of notice nor lack of a roadmap.

Yes, more and sooner communication would be better, but that's not always possible. And for those who do purchase PC hardware for dedicated folding, it's in your best interests to stay well informed. Subscribe to the Announcements section of this forum. Join the beta team and Subscribe in there. Subscribe to the RSS feed of the News Blog, etc.

@ PG - I don't understand the 2 step change this time. It wasn't a 2 stepper last time. If a change must be made, just make the one change in April. 2 changes just confuses things more. You really going to readjust the bonus points formula twice? I doubt it. Why double the work?

[tear]

...

6 months are not too much to ask, are they?

No. Neither is this...

1. Anyone building a borderline BA system (16 cores or slightly more) in the last 6 months knowing that core counts get raised from time to time, and that core counts haven't been changed in quite a longer time, was short sighted or given bad advice. I certainly didn't build any GTX 480 GPU farms on the cheap in the last 6 months, did you? And I certainly didn't build any AMD GPU farms knowing that NV JIT would double NV GPU performance at some point in the future, did you?

"From time to time" (aka chaos and lack of coordination) is exactly why people walk away from FAH.
If you demand new folders to learn upfront everything about bigadv, go back in history and then guess
(yes, guess) what may possibly happen then I have no further comment.

2. People assumed incorrectly that all BA-16 work units immediately stop getting assigned 2 months from now. While not yet confirmed, things will likely happen that same as the last BA change (BA-8 WUs were available for *many* months after the last change). You got 2 months notice, plus however long the BA-16 WUs last, plus a 2 year heads up, plus a strong reminder a year ago.

Parse error. Do you know something I don't? Please share.

Slice that any way you want, but that is not a lack of notice nor lack of a roadmap.

Past practice is neither a notice nor a roadmap.

Yes, more and sooner communication would be better, but that's not always possible. And for those who do purchase PC hardware for dedicated folding, it's in your best interests to stay well informed. Subscribe to the Announcements section of this forum. Join the beta team and Subscribe in there. Subscribe to the RSS feed of the News Blog, etc.

You got directions mixed up. It's project's responsibility to reach out to donors,
not the other way around.

Wake up, other projects do it and do it better.

[mdk777]

Yes, more and sooner communication would be better, but that's not always possible. And for those who do purchase PC hardware for dedicated folding, it's in your best interests to stay well informed. Subscribe to the Announcements section of this forum. Join the beta team and Subscribe in there. Subscribe to the RSS feed of the News Blog, etc.

Yes, as I mentioned...I started warning people several months ago not to build 2p rigs in this forum.

however,

1. No one...NO ONE...echoed or reinforced my warnings.
2.while I tried to find any recent update on the project,any official announcement on its status or pending changes in the works...I could not.
hence it was just my hunch, my experience and gut feeling that led me to make the warning.

Consequently, I don't think I was able to make any impact.

The question everyone is asking:

WHY DO YOU DEFEND AN OBVIOUSLY FAILED COMMUNICATION STANDARD?
In a thread that has moved on to a discussion on the best means to IMPROVE COMMUNICATION in the future.

[Rattledagger]

1. Anyone building a borderline BA system (16 cores or slightly more) in the last 6 months knowing that core counts get raised from time to time,

Uhm, maybe I've overlooked something, but the last news-post about BigAdv, except the one last week, seems to have been made in February 2012, and nothing on these forums under announcements nor the BigAdv-section seems to indicate any changes to #cores afterwards. So for anyone joining FAH the last 18 months, how are they supposted to know core-counts gets raised ocassionally?

[7im]

You say that communications failed, and yet you knew not to build 2P systems. Please explain the contradiction.

People are saying... No warning.
I say 2 months warning, plus an undetermined open window after that to keep folding

People are saying ... No Roadmap.
I say a Roadmap set 2 years ago: BA-8 to BA-16 to what? (Yep, BA-32)

It's all a matter of perception.

Please stop embelishing on the problems that could be managed better. (No, I am not denying problems exist.) State a fair and accurate description of the complaint, and it will get taken seriously. If you post emotional hyperbole, or say there was no warning when ther was or claim that everyone will stop folding when perhaps a few will, and imply that the world is coming to an end, you won't be taken seriously.

Get taken seriously.

[Grandpa_01]

The question everyone is asking:

WHY DO YOU DEFEND AN OBVIOUSLY FAILED COMMUNICATION STANDARD?
In a thread that has moved on to a discussion on the best means to IMPROVE COMMUNICATION in the future.

Sorry I just could not keep from

[DocJonz]

Off-topic reply to an off-topic aside:

* I should add that I am not affected by this particular announcement, but I can imagine the frustation of those who are. (The effect of the current dearth of Core17 WU's, on the the other hand, ...... )

viewtopic.php?f=16&t=25438&p=254162#p254162

@bruce - if you noticed; (a) On page 1 of that thread I already made an observation regarding WU assignment, so was reading/following the thread, (b) my comment in this thread was somewhat tongue-in-cheek (hence the phraseology and wink).

I don't really think it is that far off topic, as this thread is also discussing communication issues - was there an announcement about Core17 drying up? - no, I don't think so, at least not until donors pointed it out and sought clarification ....

[bruce]

Core17 WUs drying up is/was a matter of supply and demand. The Pande Group tries to keep enough servers with enough WUs to satisfy everyone's demands but they are at the mercy of unpredictable demands. Once the supply drops below a critical level, it takes both time and effort to respond to that shortage. You can't pre-announce unpredictable events. (I think Dilbert had a cartoon about the pointy-haired Boss demanding to know all of the unpredictable events for his project and when they would occur some time back.)

Then, too, Stanford is on holiday break with reduced staff.

New projects on new servers have to go through several testing phases before they can be available to meet those needs. See a recent post from VijayPande.

On-topic posts are about "Change in BA requirements" There's another topic about Core17 shortages. Why should I have to tell you your question has already been answered here?