bill1024
From what I am seeing the SMP points are low. Way low.
If they need them done, and done quickly then they should increase the K factor (multiplier) some how.
I still think a sliding scale Faster return better multiplier. (if they can program it)
To the point where it is an incentive to build and use the most powerful you can afford, or want to build.
I agree there is nor reason as far as I can see that it should not be at least a heck of a lot closer.
 
I thought the - work =pay was because the GPU produce more TFLOP of work that's being done because they are more powerful than a CPU.
Is that correct or not?

In a way yes, simplified a GPU will take a protein from point A to point B faster than a CPU will so the 40,000 GPU atom is actually further along in the folding process than the 160,000 CPU atom how much further I have no Idea. At least that is the way I understand it to be.  And as far as the smp issue goes when Stanford needs them done because of a backlog or anything else for that matter they will raise the points on them, (they know how to get them done) until that happens I see no reason to think otherwise.
 
I am sorry to say that you kind of got played over at the FF and that is something that should not have happened and it kind of pisses me off. The people over at the FF with Blue names are PG members, Red, Orange and Green are FF Staff and have no association with PG other than the FF. Sometimes they like to act like they have special insider information but I seriously doubt it, they are wrong too much of the time.

Grandpa_01...It is not = work for = pay,  it is = pay for = science.  if it was = work for = pay then smp would have a higer value than GPU the largest GPU WU is 40,000 atoms the largest SMP is 167,000 atoms the bigadv are 1,340,422 atoms now you can figure out the wattage and investment it takes to complete each if you wish but I am pretty sure you do not want a = work for = pay scenario because GPU would go way down then there would be a whole other group of  really pissed off people...

FYI, the largest GPU Project is 8902 (available on the IRC) which contains 75,000 atoms. Moreover, judging a Project's difficultly solely on atom count is part of the picture. Explicit (water molecules interact with other atoms) is tougher than implicit (water molecules are in the "background"). Also, the "atom density" (how many atoms in a given calculation box) is an important factor to consider. There might be other factors too, but that is all I am aware of.
 
Proteener did state that longer trajectories are better for science. One measurement that I know which might be used is ns/day but am not sure if that represents a true picture across the board (GPUs and CPUs) with FAHBench (It hasn't been updated with OpenMM 5.1).

bill1024

BadBertie
I'm still working my way through the comments here, so sorry if I repeat previous posts,  but would it not be possible to link a second 24 core board from another to make a 48 core machine? I seem to recall reading something along those lines when I was investigating folding 12 months ago. Whether you can do it with all Supermicro/Ubuntu boards/OS with dual Lan is another matter.  Did anyone look at that possibility at all?

No that is called a cluster and FAH will not work on a cluster.

Thanks Bill. Maybe when Gromacs do a re-write .........

 

texinga
....
Unless I hear it from PG (direct), .....
.  I'm still watching to see what they do before I would consider investing much of my time, equipment and hard dollars in Folding.

BadBertie...Thanks Bill. Maybe when Gromacs do a re-write .........

AFAIK, GROMACS does support Clustering, it is just that F@H has chosen not to implement/use it in their FahCores. Highly doubt that it will appear in the near future since there isn't a significant demand for it and also for the reasons that Punchy stated. Occasionally, if someone asks about clustering support, we tell them to run a single instance of F@H on each node.

Afterburner

sbinh

 

texinga
....
Unless I hear it from PG (direct), .....
.  I'm still watching to see what they do before I would consider investing much of my time, equipment and hard dollars in Folding.

This...

+1

Grandpa_01...Panther-X we both know the 8902 is not available to the general public also it has some problems. GPU folding may be the future but the present is smp and is not changing very fast. Also allot of us donors are not scientist we are (John Q Public) and also we are the majority of the donors. My mind does not work like a scientific mind everything I know says 167,000 is greater than 40,000 now we all know it cost us less to run a 40,000 atom WU on a GPU and we get paid more, so to alot of us that equals a greater value than running a smp on a CPU .Personally I have X budget to donate, so just like any other person I am going to look for the best deal I can get which is GPU and bigadv. If the scientist wish to get there work done by (John Q Public) they are going to need to think like him not like a scientist. It is actually pretty simple (perhaps too simple)
 
By the way as far as the difficulties of any given type of protein folding goes it does not matter to me what matters is the bottom line in other words points = science value in my mind.

Absolutely correct! An Average Joe need not understand all the hardware/software complexities to decide what hardware set-up to fold on. if you don't provide the right incentives to a particular set-up, eventually, it will be forgotten. I just hope that this issue is addressed sooner than later in an easy-to-understand-to-an-average-Joe manner

Eh.. my stupid rant at FF:
 
I sense they still circle the wagons, point to another illusion and then go forth and blame other things.  Things other than poor decision making by the FAH folks.  Really?  Bitcoin mining was brought up as a method to supplement income for folding.  I.E. if I make enough I can justify electric costs.  Another one of those 'visit other forums and stop living in your bubble' messages that are being presented here.
Face it... FAH has done a lot to shoot itself in the foot and it's bleeding out because no one wants to act in a reasonable and responsible manner.
I've seen two folks here that would rather defend the status quo and hope their boot licking works rather than institute changes needed to keep this project alive.
I've posted this on my forum in case it gets ummm... lost.  :)  Good day.

 bruce » Fri Dec 27, 2013 3:49 pm

Viper97 wrote:I sense they still circle the wagons, point to another illusion and then go forth and blame other things. Things other than poor decision making by the FAH folks. Really? Bitcoin mining was brought up as a method to supplement income for folding. I.E. if I make enough I can justify electric costs. Another one of those 'visit other forums and stop living in your bubble' messages that are being presented here.

Face it... FAH has done a lot to shoot itself in the foot and it's bleeding out because no one wants to act in a reasonable and responsible manner.

I've seen two folks here that would rather defend the status quo and hope their boot licking works rather than help present changes needed to keep this project alive.

I've posted this on my forum in case it gets ummm... lost. Good day.

I disagree -- and it has nothing to do with boot-licking. It's just a rational statement of possibilities that you are not considering. As I just said, their decisions are based on maximizing FAH's total science output, not on maximizing the status of specific individuals or teams. I fully understand your perspective but when it happens to conflict with FAH's scientific goals, there are many options which don't happen to match what you call reasonable and responsible.
bruceSite Admin Posts: 16012Joined: Thu Nov 29, 2007 2:13 pmLocation: So. Cal.Private messageTop

---

Edit postDelete postReport this postReply with quote

Re: Change in BA requirements

by Viper97 » Fri Dec 27, 2013 4:43 pm
Just a quick question Bruce...

How does one rationalize possibilities? Isn't that irrational in itself? Possibilities, probabilities, timelines cannot be rationalized.

Then again I'm a simple man.

As I pointed out last year Vijay has too many jobs - shoes to fill
 
he can't be teacher/prof/ head researcher/head development team/ PR/Business leader at the same time
there are other research project he has to lead - (some unrelated to FAH)
he has teaching
he has mentoring
he has his own research - new and old to continue (again not necessarily FAH related)
His own PR - let alone FAH or any other research
funding for multiple research projects (for his projects as well as staff)
dealing with Stanford board
Software development team
other research(he may be dev of) that is already at business level - that may or may not be Stanford related
and still do seminars and conferences?
 
which is why I stated it needs to operate like a business
they need a Business administrator - who can act and manage it as a biz
 
look stanford runs as a business(ot's a school that relies on funding) - FAH should be one aspect of that business - operate it as such
it brings in research grants and gov funding - business donations and other funding as well
that is stanfords revenue - it's blood and guts - without it they reduce staff and research
 
vijay has to pick a role and stick to one 
or be the administrator and give up the other roles
 
Im not beating up Vijay -he's done a great job 
 
but it's time for it to grow and operate like the business it is
it;s time for him to either be head of it and resign some roles or go back to doing research
put someone who is better at managing big business and operations as the head of the project
 
the fact you are losing folders is a sign of mis-management

bobc36
Welp, thats a sucky announcement. I was hoping they would add more project to the -bigadv pool instead of cutting the pc's capable.  I am willing to sell my G34 4p systems if someone would make a reasonable offer. 

Bobc36, If you still have your G34 systems for sale, please get in touch with me!
Marty
my email is my username @ hot mail dot com

Re: Change in BA requirementsby troy8d » Sat Jan 04, 2014 11:39 pm
It appears that most everybody can agree that the decrease in points of 50% to 75% when switching from bigadv to smp is both disheartening and unpalatable to those affected by it. Further, it appears this far exceeds the design parameters of the bigadv program Dr. Kasson outlines, more than three times the intended 16.7% decline in points when switching from bigadv to smp WUs.

My initial hunch was that, due to the exponentially decreasing QRB, the large decrease in points was a failure of points system to properly account for the speed of a bigadv machine folding smp WUs much faster than bigadv WUs. A few quick back of the envelope calculations show that a bigadv machine getting 200,000 PPD on a smp WU is completing WUs 4.7 times faster than an "average" smp machine that gets 20,000 PPD on the same WU is only receives a QRB that is 2.2 times higher. The comparison between a 10,000 PPD and a 20,000 PPD machine is quite different. The faster machine is completing WUs 1.6 times faster and receives a QRB multiplier that is 1.3 times greater. It is readily apparent that additional speed increases on slow machines are rewarded much more than equal increases on fast machines. Return speed of WUs is initially highly valued, but adding additional return speed is subject to exponentially diminishing marginal returns..

Comparing bigadv vs smp on the same machine, however, provides an interesting result. The extremely large k-factor for bigadv WUs creates a roughly equal QRB multiplier for bigadv and smp WUs on the same machine across a wide range of WUs. If the k-factor is intentionally calculated to achieve this result (and consistency across a range of WUs leads me to believe it is intentional), the reason for the large decline in points when transitioning from bigadv to smp would not lie in the QRB. This leaves the assigned point as the most probable cause for the large deviation for the anticipated 17% decline in points. This is further supported by what Dr. Kasson indicates above, there are scaling issues both ways on bigadv vs smp (bigadv WUs fold much faster on a bigadv machine relative to the benchmark machine while the opposite is true for smp).

A potentially novel solution to the benchmark scaling issue may be to simply change the benchmark machine to a bigadv class machine. This would minimize the scaling issues (both up and down) for big advantage machines and provide stability through the bigadv to smp WU transition. Additionally, as a result of this change we can expect a bit of a point boost on non-bigadv machines running smp WUs that do not scale well to a large number of cores. Most people have argued that these lower smp WUs are under-valued and under-appreciated in regard to the points they receive, potentially making the point boost an ancillary benefit rather than a detriment. Regardless, at a minimum it would simplify the scaling issues by reducing them to one direction.

If this does not resolve the benchmarking issues, I fear the solution becomes a bit more complicated. The long-term solution may require that benchmarking procedures be changed and the golden standard of a single benchmark machine for all WUs be abandoned as the real world results indicate that it is not rewarding "equal pay for equal work." (I use the term work as a measurement of science). For bigadv class machines, the aforementioned scaling issues and observed data points suggest that, on a bigadv machine the bigadv points are over-inflated and the smp points are under-inflated relative to the theoretical construct of the points system. It would be extremely disheartening if PG chose to sacrifice the integrity of the observed results of the points system to maintain its underlying theoretical integrity.

In attempting to break away from the strict adherence to "equal pay for equal work" it is important to note that it appears bigadv was never based purely upon the premise. It is clear both from Dr. Kasson's description of bigadv and the observed comparison between bigadv and smp that the bigadv points premium rewards factors other than pure work/science, negating the direct application of equal pay for equal work to big adv WUs.

Given that actual results of bigadv vs smp PPD manifest large deviations from the intended behavior, the aforementioned difficulty in benchmarking bigadv WUs using current procedures and that equal pay for equal work only loosely applies to bigadv WUs; it may be time to consider revising benchmarking procedures to produce more consistent results. Several years ago, GPU WUs were benchmarked separately from smp WUs because it was not a valid comparison (and infeasible at the time). The description of how bigadv WUs are benchmarked makes it very much appear as if the integrity of the points system is being compromised for the sake of a maintaining its theoretical purity.

Other potential approaches to addressing the large decline in points transitioning from bigadv to smp can be achieved by modifying the points system. If benchmarking is at the heart of the problem, however, this approach will be a second-best solution. One possible approach is to alter the points formula to punish increased return speed less harshly than the square root function does (outlined above). I've briefly played around with a few functional forms that do so without drastically altering the points landscape. This approach would, however, reward increased points to all smp folders as well as QRB GPU folders. An alternate approach would be to create a smalladv WU type with a more appropriate decline in points relative to the bigadv level. Smalladv could be similar to bigadv but targeted at some of the larger SMP WUs that scale well with more cores and would benefit from the increased folding horsepower that no longer qualifies for bigadv.

Regardless of the cause, it is readily apparent that the bigadv point system is not functioning within the design parameters Dr. Kasson described, and this has understandably upset a large number of donors that will be negatively affected by the upcoming changes. Any changes to benchmarking procedures or points system will have to be carefully considered and probably will not be in place by the announced deadline for upcoming bigadv changes. As a show of good faith, it would be a nice gesture if PG delayed the impending changes until a solution is worked out or a temporary stop-gap measure can be put in place. This would prevent affected donors from experiencing a massive decline in points that it appears they were never intended to face if bigadv points functioned as originally outlined. While it is important that the folding community embraces change and continues to move forward, it is equally important to fix the issues at hand before doing so. The biennial cycle of bigadv crises has quickly grown wearisome.

troy8d Posts: 1Joined: Thu Oct 25, 2012 7:05 pmTop

It makes me wonder why they are even still using points in the manner they are...
 
Take a few moments and look at the production of the top five teams... All I can say is hmm... http://folding.extremeoverclocking.com/team_list.php?s=&t=33#33
 
And good work Troy!!!

I see Kasson posted a reply:- https://foldingforum.org/viewtopic.php?f=16&t=25411&start=405#p255302
 
Roughly translated, and to save everyone the hassle of visiting FF it goes something like this.
 
Many thanks for listing all the problems and some ideas for solving them.  We are going to carry on with the changes as we previously stated in December.

Well, I was hoping for a different response, and at the same time was not very surprised. 
 
Accelerating the date to eliminate lower-end Bigadv rigs...guess they must have figured "well whada we got to lose"!