DAB & transparency

[rjbelans]

7im is correct, we don't make the decisions in the DAB forum. We have the open discussions about the topics and then PG makes the decision they feel is appropriate based on the feedback we have provided. He knows this because that's what was posted in the FahBlog when the group was formed.

I've reread mine and mdk777 post and do not see where either of us thought that decisions were made in DAB. So I don't know were 7im got that idea.

I didn't know what you were questioning about 7im's post when you said "How do you know that?". I thought I was acurate in responding to your question, but it seems you were questioning a specific part of his post (as seen in your next response) and not the whole thing.

We discuss things, like the QRB for GPU, changes to BA, and opening the Beta forums to public viewing, in there in advance of a decision being made and announcing it publicly.

Then things are being discussed that may and have impacted donors.

We discuss our thoughts on what PG has posted as possible ideas. The things discussed may impact donors, or they may just become ideas that never get implemented. As I said, PG uses their judgement based on our feedback on the topics being discussed.

If everything in there was publicly viewable, then there would be no use of even having it because eveybody would be posting about the topics outside of the DAB forum section (probably more on other forums).

And the mods would be locking post like that as they did when people were posting about beta projects in the open forum.

I would think that as a DAB representative would want as much feedback from their team on things being discussed that could impact them.

I do get feedback from my teammates, all the time. I try to read every thread on my team page every day. Based on what is posted, I am able to compile a collective response on the topic at hand. Just because they are not directly aware of the topic of discussion doesn't mean that they are unable to provide me with valuable input. I believe my ability to pull this kind of information together is one of the main reasons why my teammates voted for me to represent them on the DAB.

[orion]

rjbelans, I do appreciate the time that you do put into the project as do I anyone who donates time, hardware and money. I don't visit your forum so I’m not familiar with the flow of information there.

As a donor I would like to read what is being discussed on my behalf that may or may not impact my folding.

[mdk777]

Not sure what points I'm proving by letting you know that I didn't have the reference on hand to show you where Dr. Pande said that QRB for GPUs was in the plan, but, now that I am home from work and have time to look it up, here it is (viewtopic.php?f=16&t=19042&p=190618#p190618).

And then, the other point I am proving is? That Dr. Pande gave plenty of notice that this was in the plans?

1. Rapidly...time delays

I also was aware that it had been mentioned in the past. (1.5 years ago apparently)

Setting up a shared storage system was also mentioned.

I participated in the BETA.

However, when you need to go back in time to dig up references to what might happen, or might not happen,... well that is not effective communication is it?

It might as well be communicated via rumor.

I mentioned to my wife 20 years ago that I would like to live in Alaska.
Knowing this will not lessen the shock when she finds the house sold and a container moving out the driveway bound for the Yukon.

I understand you don't want your considerable donation of time and consideration criticized.
I am not. I am explaining that the average folder is not getting timely and accurate information.
The structure of the project, (not your participation in it) is systematically flawed.
The premise that information is given on a need to know basis, after the fact (decision already made) is not one designed to establish trust and confidence.

[Grandpa_01]

Any time there is a secretive society there is going to be problems, Yes PG did say they were considering a GPU QRB bonus and if they would have given an explanation of what that would mean to folding it would have been better but there was no hit at all as to what the implications were. There most likely would have been some different builds going on if people knew which way folding was headed and what the rewards would be for what was being worked on at the time. We still do not know if they can port all of the WU's over to GPU or not and what the implications of that will be. Information is invaluable when it comes to choices of what to do when upgrading and without it there is always going to be problems. It does not take that much to say this is what we are working on and this is what it looks like at this time.

I believe that at the time the statement about the QRB bonus was mentioned so was the statement about PG wanting to have a greater transparency, well the QRB is coming to light but the transparency is still very dim almost non existent. The DAB is great for those involved in it but for those who are not it is basically worthless we know no more than we ever did and that is not being transparent. The donor participation has dropped at an alarming rate recently which says something is not working, you can blame the economy or anything else you want but other projects are not seeing the same thing, why is that does that not pretty much rule out the economy.

Could it be the lack of donor support, poor communication, those are the most common things a person will read on other folding forums when they are disgruntle about folding it is a common thread among donors. Anyway what is the answer I do not know for sure but I think if it were me I would start with those 2 items, as far as the points system goes I believe it is fine, it works to point people in the right direction just a little more advanced information including the implications of the changes that are being considered would be helpful to people considering there upgrades.

You are never going to be able to satisfy all the people but you certainly can attempt to lessen the affect of changes by getting the information out there, especially when something is going to have a very large affect on the folding community as a whole.

[7im]

Any time there is a secretive society there is going to be problems, Yes PG did say they were considering a GPU QRB bonus and if they would have given an explanation of what that would mean to folding it would have been better but there was no hit at all as to what the implications were. There most likely would have been some different builds going on if people knew which way folding was headed and what the rewards would be for what was being worked on at the time. We still do not know if they can port all of the WU's over to GPU or not and what the implications of that will be. Information is invaluable when it comes to choices of what to do when upgrading and without it there is always going to be problems. It does not take that much to say this is what we are working on and this is what it looks like at this time.

I believe that at the time the statement about the QRB bonus was mentioned so was the statement about PG wanting to have a greater transparency, well the QRB is coming to light but the transparency is still very dim almost non existent. The DAB is great for those involved in it but for those who are not it is basically worthless we know no more than we ever did and that is not being transparent. The donor participation has dropped at an alarming rate recently which says something is not working, you can blame the economy or anything else you want but other projects are not seeing the same thing, why is that does that not pretty much rule out the economy.

Could it be the lack of donor support, poor communication, those are the most common things a person will read on other folding forums when they are disgruntle about folding it is a common thread among donors. Anyway what is the answer I do not know for sure but I think if it were me I would start with those 2 items, as far as the points system goes I believe it is fine, it works to point people in the right direction just a little more advanced information including the implications of the changes that are being considered would be helpful to people considering there upgrades.

You are never going to be able to satisfy all the people but you certainly can attempt to lessen the affect of changes by getting the information out there, especially when something is going to have a very large affect on the folding community as a whole.

We are all involved in the DAB. Anyone who posts in the forum is heard, and represented by Bruce in the DAB. So it is not worthless by any means. If you think the DAB is worthless, please ask Kendrak to resign and pass your team's seat on to the next team.

[derrickmcc]

Not sure what points I'm proving by letting you know that I didn't have the reference on hand to show you where Dr. Pande said that QRB for GPUs was in the plan, but, now that I am home from work and have time to look it up, here it is (viewtopic.php?f=16&t=19042&p=190618#p190618).
And then, the other point I am proving is? That Dr. Pande gave plenty of notice that this was in the plans?

I also was aware that it had been mentioned in the past. (1.5 years ago apparently)
...
However, when you need to go back in time to dig up references to what might happen, or might not happen,... well that is not effective communication is it?
...I am explaining that the average folder is not getting timely and accurate information.
The structure of the project, (not your participation in it) is systematically flawed.
The premise that information is given on a need to know basis, after the fact (decision already made) is not one designed to establish trust and confidence.

So:
1) PG posted on July 3rd 2011 that they are considering QRB for GPUs;
2) PG posted on 21st October 2012 on the blog:unified-gpusmp-benchmarking-scheme-equal-points-for-equal-work that they are going to test QRB for GPUs by releasing a Beta Test WU;
3) 27th October 2012 Project 8057 is released for Beta Test.

Given that QRB for GPUs is still in Beta, at what point exactly is PG meant to have informed us (again) that they were intending to do this? If I understand your argument correctly, July 3rd 2011 was too early and 21st October 2012 was too late?

By the way, another blog post may be of interest: http://folding.typepad.com/news/2012/11 ... n-fah.html

[orion]

If PG truly wants to be transparent they would open up DAB to viewing.

Knowledge is power, are they afraid of donors getting little knowledge or insight into their possible donating future...I hope not.

[Macaholic]

Given that QRB for GPUs is still in Beta, at what point exactly is PG meant to have informed us (again) that they were intending to do this? If I understand your argument correctly, July 3rd 2011 was too early and 21st October 2012 was too late?

Actually, it was first announced May 29th, 2008;

GPU2 points & benchmark machine (and points in general)

I've talked about this in the Folding Community Forum (FCF), but I thought it's worth reminding people (especially NVIDIA-based folders new to the project). As the GPU2 code matures, we may need to change the points per day (PPD) and/or the nature of the benchmark machine. The PPD for the bechmark machine may go up or may go down depending on what change we decide in the end. The main problem is that both the NVIDIA and ATI code is undergoing optimizations, which means PPD will change in complex ways. For example, if the NVIDIA code doesn't change, but the ATI code gets optimized and our benchmark is based on an ATI card, then NVIDIA PPD will go down, even if the NVIDIA code doesn't change.

Also, how efficient these cards are depends on the protein we run on a given card. To address this issue, we will likely switch the benchmark machine to run a card whose PPD does not change so variably with protein, although that may be hard to find (we'll see).

Anyway, just a heads up -- the PPD we see today is not necessarily what you'll see later. Your PPD may increase or decrease depending on many complex factors (code optimization, benchmark machine, protein simulated, etc), and I expect there may be more points variability with GPUs than with other platforms (isn't life on the bleeding edge fun). However, due to the great performance of the GPU's right now (which will likely only get better), the GPU PPD would be significantly greater than a typical CPU client.

I should also remind donors that we've been talking about rebalancing points in general, to try to get SMP, GPU, PS3, and classic points in line with the scientific productivity done by the clients. As long as more points = more science, donors can optimize for points and that will automatically optimize for maximum scientific output.

[Macaholic]

If PG truly wants to be transparent they would open up DAB to viewing.

Knowledge is power, are they afraid of donors getting little knowledge or insight into their possible donating future...I hope not.

Everyone is entitled to their own opinion. Knowledge is power, and Dr. Pande does give donors knowledge and insight via his blog and the forums here;

How FAH works: Folding on streaming processors (GPU's and PS3)



One of the critical issues in computer science right now is the limits to how fast a single CPU can calculate. While Moore's law is still going strong in a literal sense -- i.e. the number of transistors which one can put on a chip is doubling every 2 years or so -- this doubling of transistors is not leading to a doubling in CPU speed as it used to over the last few decades. Well, at least not for typical programs (eg Microsoft Word). In order to get big speed increases, there'sa major change in the programming paradigm. One key change is the existence of "streaming processors." GPU's and the Cell processor in the PS3 are both examples of streaming processors.

What makes streaming processors potentially much faster than regular CPU's is how they handle computation vs memory access. Normal CPU's use lots and lots and lots of transistors on cache (local memory on the CPU chip to help keep the CPU fed with data and instructions). Streaming processors use the additional transistors on additional computing elements (eg floating point units). By doing so, they can do lots of FLoating point OPerations per second (FLOPS) in an optimal situation, although getting one's code to behave optimally is not easy. Typically this means balancing FLOPS with memory access to make sure that there's data available for calculation. This has been the primary challenge in our GPU and PS3 codes, and is something which we have, for the most part, figured out for a significant subset of the calculations we run on FAH.

These advances have lead to our GPU and PS3 clients. The family history of all of this starts with the GPU core. This GPU code was then brought over to the PS3 and enhanced. We are working to bring back some of those scientific enhancements back to the GPU code. This is all pretty bleeding edge, but so was distributed computing in 2000 when we started. Our expectation is that given how modern processors are developing, in 8-10 years streaming processors will be much more standard and will be a major way in which FAH works.

Finally, it's also interesting to think of how CPU's may themselves turn into streaming processors. As CPU's add more cores, they start to have more functionality similar to streaming processors in a limited way. Perhaps more interesting are some of the new chips rumored to be developed at Intel and AMD/ATI. Intel's 80-core chip is very interesting and something which our code would likely run well on. Also, the fusion of AMD's CPU's with ATI's GPU's could be very exciting, potentially bringing the best of both worlds. We're looking forward to these and lots of other emerging technology. FAH is running very fast now (over a petaflop, i.e. 1,000,000,000,000,000 floating point operations per second !) and we look forward to continuing to push the frontiers.

[mdk777]

Given that QRB for GPUs is still in Beta, at what point exactly is PG meant to have informed us (again) that they were intending to do this? If I understand your argument correctly, July 3rd 2011 was too early and 21st October 2012 was too late?

So you are really defending an obscure single line post buried in a thread from a year and a half ago as adequate notice to the FOLDING community at large?
A post that actually says that while it is under consideration, it is not feasible at this time?

Yeah, something between the two posts updating on the process under consideration might be a good idea.

Something a month after the launch of a Beta, discussing the success, or limits, or ongoing issues needing to be resolved might be a good idea.
YES, that is exactly what people are asking for in about 100 posts so far:

1. will all GPU projects be converted to QRB?
2. can SMP projects be converted to GPU if folders decline to run SMP in favor of GPU?
3 are the points currently reported on the beta a function of a small test unit, or are they representative of the new gromacs?
can we expect future GPU projects to have the same level of responsiveness ( lack of screen lag) or is this a fluke of the test unit?

I am really amazed that we are even having this discussion. The entire reason these threads exist is because major donors... say like Grampa_01....voice concern that they can't figure out where things are going. To turn around and defend random posts every 1.5 years is the ultimate in absurdness.

Of course a 1.5 year gap in communication is absurd.

According to your logic, we should not expect answers to the above questions in what? Six months to two years?

REALLY? You needed to ask if this is a good idea? You need to ask if this is what donors are looking for?

Everyone is entitled to their own opinion. Knowledge is power, and Dr. Pande does give donors knowledge and insight via his blog and the forums here;

Really? A general post from 2007 is what donors should use to guide them in their daily decisions. Now we have gone from updating donors every 1.5 years to explaining trends and goals every 5 years. No reason anyone could ask for anything more.

[mdk777]

Actually, it was first announced May 29th, 2008;

EXACTLY

You know what? When people don't see something happen for 4 years, they have this odd habit of assuming that it is no longer expected to happen.

If you are still intending to do something...you might want to give an update, a little heads up....say a six month warning when the feasibility of doing it again comes under discussion ... and a three month warning that software looks like it is progressing to make a beta to test the process a reality...and then a monthly update on how the beta and ultimate implantation is progressing. You really think that this level of communication is onerous or unfeasible?

Your posts just continue to prove the complete ALICE IN WONDERLAND logic that is the culture of this site(posts every 4 or 5 years are adequate), and which continues to alienate donors in huge numbers who have become accustomed in their lives to a slightly higher level of customer service.

[orion]

If PG truly wants to be transparent they would open up DAB to viewing.

Knowledge is power, are they afraid of donors getting little knowledge or insight into their possible donating future...I hope not.

Everyone is entitled to their own opinion. Knowledge is power, and Dr. Pande does give donors knowledge and insight via his blog and the forums here;

5 year old blog post is a bit dated isn't it?

It my still hold true over all but what does it have to do with DAB, DAB wasn't in existence then and my not have even been thought of at that time.

Still the point is that what is being discussed in DAB now, may or may not impact us as donors and we will not know until it is decided on by PG and made public.

By the time that we do find what PG has decided our hardware purchases may have gone in the wrong direction. Some of us don't have that kind of money to throw around.

[JimF]

By the time that we do find what PG has decided our hardware purchases may have gone in the wrong direction. Some of us don't have that kind of money to throw around.

Yes, but PG doesn't know what decisions they will make either until they make them. That is research; it isn't a Ford production line. (And if they decide that we should be part of the decision on what research should be done to "protect" the older hardware points, I am bailing out at that point.)

On the other hand, a new computational technique may come along tomorrow which blows everything we have done thus far out of the water. Those are the chances you take for science.

viewtopic.php?f=38&t=17469&p=174143&#p174143

[mdk777]

Yes, but PG doesn't know what decisions they will make either until they make them. That is research; it isn't a Ford production line. (And if they decide that we should be part of the decision on what research should be done to "protect" the older hardware points, I am bailing out at that point.)

Total reversal of everything said in this thread.

Others have indicated that this was in the works for 5 years and that we should have expected it.
You say that the vagaries of research change overnight.

You both can't be right. But perhaps long established goals suddenly become feasible.

However, communication improvements would address both. In fact, if you are correct, prompt communication becomes all the more critical.

"we understand that we had said this is the direction...but here are the reasons we are making a sudden departure from that course:...."


Finally, in case anyone was wondering, I have always supported the QRB, and have always supported giving the GPU points equal to the work done.

Heck...I pulled up a post from 7 years ago of mine proving my consistent position on the topic.

The point of this tread was not to hinder the progress of science, but to point out that better communication with donors would be an expedient.
I actually wholeheartedly support the current beta.

However, defending the way it is being communicated is the biggest joke I have ever read in life.
To those who see no flaws in this system, you are only deluding yourself.
The vast majority of folders who don't post here have already voted with their feet.

[JimF]

They can only communicate what they know. As you point out, they have been communicating for a long time the direction they are going. Part of the problem is that as a research institution, they don't want to get into recommending one brand of GPU verses another, or even say that "CUDA is better than OpenCL" (or vice-versa) too blatantly. They would get into a uncomfortable position with the companies they need to work with, among other problems.

But no matter how you make the decision or communicate it, someone will always be on the margin who just bought a new system that they wouldn't have otherwise. Changing the communication, even if that is needed, will only shift the decision point a little earlier or later, and change the people left out in the cold. So my advice has always been to not over-buy on a given technology. If someone wants to sink a lot of money into a given hardware level, then (as with the housing market), they could do either well or badly. You will hear from the ones who do badly.