Folding Forum

k1wi wrote:They have (when they outlined the QRB)

Fair enough, that's before my time. Or maybe I didn't search hard enough

HaloJones wrote:With regards to points inflation I echo the comments that we already have it. We have got to stop looking backwards, constantly worrying about "old-timers". I'm one of them having been in the original genome@home program. Personally, I'd be in favour of resetting the clock, declaring the current points race finished, publishing them somewhere permanent and moving on to Folding@Homev2 with all new rationalised points across all units. QRB for all!

Small correction, it would be Folding@Home version 3, since FAH has already re-started at zero once back in...

... well, trying to find a correct date for this is very difficult, since FAH seems to have deleted all news older than 2007 and these forums blew-up for some years ago so no information here either...

So, I'm unsure if it was in 2001 or in 2002 they did re-start at zero.

As for QRB, this is one of the main reasons FAH's point-system is FUBAR, just resetting to zero again won't change this fact.

EXT64 wrote:That is there for a specific reason - they want to encourage fewer fast machines over more slow ones. Without QRB 2 computers at 0.5x speed are equal to one computer at 1x speed. But to the scientists the 1x speed computer is a little more valuable as it can process a longer Gen in a reasonable period of time and thus get a project done faster (whereas the 0.5x will require more projects and get them done slower).

Where's multiple holes in your reasoning:
1: Two slow computers will each use maybe 100 W, for a total of 200 W while one fast new computer will maybe also use 100 W meaning it's a big advantage to upgrade since this will lower the power-usage and the cost of running the computer.
2: One computer takes less space and often makes less noise.
3: Keeping one computer running is easier and takes less time than keeping two computers running.
4: Would expect majority of users upgrades their computer due to other reasons, and not exclusively due to DC-computing. Keeping the two slower computers running after the upgrade will finish more wu/day than only running the new computer.
5: Anyone with two slow computers and not the funds (or reason) to upgrade immediately will often not use the two slow computers since the point-system severely discourages running older computers.

6: Users are competive, they wants to get their computer to #1 by RAC on the top-host-list.
7: Some users wants to get to #1 within the group of users running N computers.

8: For all computers not only sitting in a corner and only crunching FAH (and dust-bunnies), the point-system severely discourages crunching on computers also used for other things.

#6 and #7 isn't an option for FAH, since FAH is currently missing the neccessary stats-info.

#1, #2 and #3 directly shows upgrading is an advantage, even without any bonus at all. This is especially true for #1.
#4 and #5 shows two examples why the point-system discourages users to crunch on all their computers.

For #8, let's give two numeric examples:
a: Computer #1, 61895 PPD if doesn't use computer. If uses computer for 1 hour/day, it drops to 53267 PPD, this means 14 % decrease even actual crunching only 4 % less. ==> I'm not crunching FAH on this computer.(*)
b: Computer #2, estimated 12500 PPD. 1 core of 6 used for 12 hours/day ==> drops to 9500 PPD since forced to only use 5 of 6 cores, a 24 % drop while actual crunching only 8 % less. ==> I've not tried to crunch FAH on this computer in 2013.

My guess is many users has usage-patterns like my a-example under #8, where computer can crunch FAH for maybe 23 hours/day but computer is used, or in reality often only 1 core of computer is used, for 1 hour/day. The current FAH-point-system severely discouraged users to use their computers, even it's only 1 hour/day, and my guess is some of these users comes to the same conclusion as I've landed on, to not run FAH at all on the affected computer.


As for "faster overall", how this work I've never understood, but let's make a small example...
Making it very simple let's say FAH has 200k active computers (it's unclear of GPU is part of this or not so I've not included these).
Let's say all computers needs 24 hours/wu, each project has 10k run/clone-combinations and let's say a serie needs 100 generations.
Let's say 50 % of the computers runs 24/7, and let's say 50 % runs 23/7 ==> 100k runs 24/7 and 100k runs 23/7 ==> 10 projects is finished after 100 days and an additional 10 projects is finished after 104 days.

Let's say 10 % of the 23/7-computers is shut-down due to users not liking the QRB-system ==> 100k runs 24/7 and 90 k runs 23/7 ==> 10 Projects is finished after 100 days and an additional 9 projects is finished after 104 days.

Uhm, so, how is finishing 19 projects in 104 days "faster overall" than finishing 20 projects in 104 days?


(*): I did try running this computer during the day, but out-of-season thunderstorms re-booted computer and GPU doesn't restart without logging-in. This is coupled with GPU crapping-out if tries to run a game.

Rattledagger wrote:Where's multiple holes in your reasoning:

A lot of generalisations in there… I'm not claiming to be typical, but I've got three (otherwise redundant) ~2GHZ Core 2 Duo machines upstairs in a spare room which crunch away quietly* 24/7, running headless they take about 130W between them. They contribute less than 6,000 PPD (~15% of my total), I don't suppose they contribute much science but more than if I left them switched off!

The displays get woken up about once a week or so to see if either Apple or Ubuntu have issued any useful updates, and that's about it.

I don't mean to dispute the gist of what you're saying, but I suspect the spread of computers, usage, motives etc etc may be wider than your post implies.





* Not so quietly if the fan bearings on one of them haven't warmed up, but that's a separate issue

billford wrote:I can see the logic in that… but (if I understand the PRCG "tree" correctly) the two 0.5x machines can be processing generations in two clones, thus exploring two trajectories.

Whether that's better (faster overall?) from the pov of the scientists only they can say…

Several comments along these lines already but the scientists have been very emphatic when they've said that two short trajectories are significantly less valuable that one trajectory that's twice as long. The whole point of bigadv is that a trajectory that goes 48 times as far in a given calendar time discovers important facts that 48 individual trajectories cannot discover. I can't say if it is 48 times as valuable or more valuable by some other factor (i.e.- how should bonus points be computed), but it's certain that without some kind of QRB, the same people running the same hardware would produce less useful science.

That is very, very different than many DC projects, where ever piece of work is independent of every other piece of work and where equal points for "equal" work means something different than FAH where processing long WUs faster is really critical.

For FAH, the definition of "equal" changes. When you compare the work done by lots of slow computers with the work done by fewer fast computers, they are NOT equal. Nevertheless, the work done by bigadv (or SMP) and work done by a GPU will be considered equal when either one can simulate the same number of nanoseconds for the same number of atoms in the same calendar time.

Personally, I'd be in favour of resetting the clock, declaring the current points race finished, publishing them somewhere permanent and moving on to Folding@Homev2 with all new rationalised points across all units.

Interesting, brainstorming can bring about ideas that otherwise wouldn't surface. Your idea is intriguing, but I don't think I can be objective in my reaction. I've been folding since 2000, and the number of work units my computers have successfully processed is a symbol to me that I've contributed faithfully for a humanitarian cause. But I'm also a long-time member of the '4P Mafia,' the majority of my production points having been realized through powerful server-class machines. OK, just my reaction without pondering this too much, and admittedly biased: I don't think I'd mind a startover with respect to points, but I would definitely want to see the total completed work units tally tied to the user rank without starting over at zero. I'd like to say that I'm all in for the good of humankind, but that would be dishonest. I enjoy being recognized for dedication and loyalty. ....just my viewpoint

Well Macaholic noted that this discussion has been going on for many years.

Vijay wrote:

I've been thinking about what we can do to help improve that in a sustainable way. We've had many attempts but keeping the communication going is probably only possible with us having someone who's primary job is donor relations and communications -- the science and development team are as it is overloaded with the tasks of keeping FAH running. So, I have been looking into hiring (into a part time role) someone who's sole job is donor relations and communications.

Just for nothing, I looked at what i wrote on the subject in 2010 and 2011

In my opinion, this part of the project is too important to be left as an afterthought, or a dreaded obligation.
While not central to the research science, it is central to maintaining the donors.

Hence it would be a good investment to have a full time person, a project manager. Someone who is not tasked with maintaining servers, or doing research papers. A person who would be both a points czar, and a liaison between the researchers and the folding donors.

With close to a half million donors, it certainly seem that a professional full time person would be justified.

viewtopic.php?f=16&t=19062&p=190802&hilit=liaison#p190802

So here is to seeing results in 2014 instead of talk.

bruce wrote:the scientists have been very emphatic when they've said that two short trajectories are significantly less valuable that one trajectory that's twice as long.

Well, I'm certainly not going to argue with them… even if I don't understand why.

Which may not be unconnected with the fact that I'm an electronics engineer (or was before I retired), not a molecular biologist

Thanks for that, and to all who made the same point.

An excellent post by Nookiebandit laying out possible communication activities/priorities for a communication director.
viewtopic.php?f=16&t=23074&hilit=dab&start=30#p230056

Again, from some years back, but the bullet points are all applicable to the discussion today.

Its not the change per see that bothers donors, but the inability to anticipate those changes in a timely fashion.

(no need for Bruce or 7im to point out news items from 4 or 6 years ago describing the possible transient nature of any project...people need to be reminded, or be able to view the information in continuous or concurrent form to understand it.)

mdk777 wrote:(no need for Bruce or 7im to point out news items from 4 or 6 years ago describing the possible transient nature of any project...people need to be reminded, or be able to view the information in continuous or concurrent form to understand it.)

It appears that from the Pande Group's perspective, giving folks a reminder 2- or 4-months before a change happens is sufficient. It appears from your perspective, that neither 4-months or 4 years notice (if that's what it was) is adequate. There's probably a happy medium somewhere between those two values, but I suspect that no matter when notice is given, it will cause disruption in people's plans.

Do I sense a rally around the castle moment? Man the towers?

It seems history here has a way of repeating itself.

What was that old quote?

OH yes...

“Those who fail to learn from history are doomed to repeat it.” Sir Winston Churchill

It appears that from the Pande Group's perspective, giving folks a reminder 2- or 4-months before a change happens is sufficient. It appears from your perspective, that neither 4-months or 4 years notice (if that's what it was) is adequate. There's probably a happy medium somewhere between those two values, but I suspect that no matter when notice is given, it will cause disruption in people's plans.

The 2 to 4 month notice you mention for the future could have been anticipated well in advance (I did and warned people asking about building 2p systems recently)
Of course, the 2 to 4 month notice for the future does nothing to help people who have recently built rigs(in the last year) and did not anticipate that any changes were in the works.

but I suspect that no matter when notice is given, it will cause disruption in people's plans.

Yeah, that is the wonderful thing about having open sources as described by Nookiebandit...people can monitor progress and the rate of change themselves and anticipate pending official announcements.
It makes the announcements themselves anti-climatic ...just because everyone can see them coming.

As discussed in the past...no one is surprised when the next version of Ubuntu launches.
If they are interested, they can follow the progress every step of the way up to launch day.

This is the insane advantage...no need to blame or feel mislead...when the information is open and continuously available you can make informed choices.

Do I sense a rally around the castle moment? Man the towers?

yeah, not trying to beat a dead horse, really

I really did think that Nookie-bandits post had some excellent points and could be used as possible bullet points.

Of course, like you imply, you really need to want to improve rather than defend.

@mdk777 Thank you for the attribution in my post from just over a year ago. I went back and re-read it, and I continue to believe those suggestions would substantially improve F@H in every subject are of interest to both casual and competitive donors.

With a year to reconsider the priorities suggested in that post, the one recommendation I believe would be the most valuable to everyone on the forum is an official Stanford F@H Communications Director. If VJ were to recruit a Junior/Senior level undergraduate student in the business school pursing a major in communications, or a graduate student in the MBA program with an engineering undergrad, I'm sure any communications student would see this as a resume-booster of the first order. Everyone would win, and it could be done virtually for free.

I'll gladly volunteer to write the job description if asked.

So what you are saying it that the work of bigadv folders is not appreciated? Are we such a nuisance to you that you need to get rid of some of us?

Leonardo wrote:but I would definitely want to see the total completed work units tally tied to the user rank without starting over at zero.

That is a fair point, and makes sense. We often ignore the number of finished WU and look only on points cumulated. But the better indicator for long term contribution is number of WU and should survive any point devaluation/reset.