Folding Forum

I'm concerned about the seemingly negative effect/reactions this has had upon quite a few existing Bigadv Folders. I don't like it either, especially going backwards (in points) with the oft complained about 2684 WUs timeframe/points. Seems a better (positive) way of dealing with the need for normalization would have been to increase non-Bigadv points to meet the desired level of balance. Taking points away from all Bigadvs (especially after years of having them in place) has an expected and predictable negative reaction from those that have invested in hardware to run Bigadvs.

Seeing people post here (and on other Folding Forums) that they will either move off FAH, reduce/sell Bigadv Folding machines or simply switch off Bigadv WU's is not a good result of this Bigadv points change. I don't recall reading a bunch of posts from people running standard SMP and GPU WU's saying they will stop Folding because Bigadv points are overall out of balance. The recent changes to the 6903/6904 points were expected and predictable, but to adjust all Bigadvs points downward 20% was not warranted IMO.

I'd bet the largest component of Folders are male and in that regard, males are "visual animals". The points system is our yardstick for progress and very important to have in place. In the 1.8yrs I've been Folding, I don't recall a large ground-swell of people saying they felt that Bigadvs were out of proportion. We had a recent poll thread right here that somehow was removed that indicated a majority of respondents were not in favor of changes to the Bigadv points. Then the 6903/6904's came out which were highly discussed/debated here and on other Folding Forums. That situation then led to the oft referred thread right here "Points are getting ridiculous" and then all Bigadv points were subsequently downgraded. That seemed to be an over-reaction to me and was not warranted based upon what I've seen from Folders opinions I've seen. Maybe the folks at FAH have some other sort of response/data, but really (from what I've seen) nobody was complaining about the Bigadv points with exception to the recent 6903/6904 Bigbeta WUs.

I just wanted to share my thoughts as someone who predominately runs Bigadvs. I'm not going to quit Folding, just wanted to give my 2-cents for what it is worth. I really hope that we do not lose Folders or reduced Folding activity over this points change because as others have said, "the science work did not change". I do question whether people really were that upset about the Bigadv points system as it was (with exception to the 6903/6904 points) and would encourage Pande to reinstate the previous standard Bigadv points. I expect that will not happen, but if people were not regularly complaining about standard Bigadv points, I'd hope that Pande would consider what actual day-to-day Folders opinions are driving these kind of changes. If it was that recent "Points are getting ridiculous" thread, then I'd say that was not reflective of what people overall wanted to have happen to all Bigadv points.

6903/6904 should have remained in beta long enough for the points issue to be flagged and altered then, they weren't and people jumped on them as fast as they could set their machine up to receive them. Then complaints break out about the points and ALL -bigadv folders were penalised not just the ones running 6903/4.

Wrong thing to do in my book, points should have been altered for 6903/only and everything else left

texinga wrote:We had a recent poll thread right here that somehow was removed that indicated a majority of respondents were not in favor of changes to the Bigadv points. Then the 6903/6904's came out which were highly discussed/debated here and on other Folding Forums. That situation then led to the oft referred thread right here "Points are getting ridiculous" and then all Bigadv points were subsequently downgraded. That seemed to be an over-reaction to me and was not warranted based upon what I've seen from Folders opinions I've seen. Maybe the folks at FAH have some other sort of response/data, but really (from what I've seen) nobody was complaining about the Bigadv points with exception to the recent 6903/6904 Bigbeta WUs.

Reminds me of Washington DC, what the people want has nothing to do with it, remove the necessary stuff and ram it down there throats and tell them its good for them...

Nathan_P wrote:6903/6904 should have remained in beta long enough for the points issue to be flagged and altered then, they weren't and people jumped on them as fast as they could set their machine up to receive them. Then complaints break out about the points and ALL -bigadv folders were penalised not just the ones running 6903/4.

Wrong thing to do in my book, points should have been altered for 6903/only and everything else left

I couldn’t agree more Nathan. Donors, such as me, that have been left with the opportunity to fold 6903/4s are not in such bad shape compared to a couple of months ago. I really do feel for the guys that, after carefully weighing up the options, have recently bought chips such as 2600Ks and are perhaps wishing they hadn't.

A key aspect to the successful implementation of any significant change is expectation setting. In this case that would have entailed, whilst selling the virtues of what's to come, discussing over time the necessary painful measures needed to achieve the end result so that when the change was made it would have been an anti-climax and not nearly as bad as expected. People would then perhaps have been more likely to accept the change with a sense of relief along with comments like 'it could've been worse'. It's regrettable that this approach wasn't taken.

As texinga alluded to, I don't think anybody would have left the community (on either side of the argument) if the implementation had been delayed, phased or stopped while people were brought onboard.

I believe you are correct that the 6903's / 6904's are the ones that should have been lowered and there were a few of the beta testers that said this from the beginning while they were in beta. In actuality you are correct also that there were not very many people griping about the points bigadv folders were receiving until the 6903 / 6904 WU's came out. In the "Points are getting ridiculous" thread the OP was referring to the 6903's / 6904's then a very vocal minority of folders started bashing the point system. Unfortunately there was very little input from bigadv folders in the thread, there was some from myself and a few others that do not care all that much about the points (They just add a little fun to folding) so I am really not a very good representing of bigadv folders. and the pole was not removed it was renamed but no one new where it was or what it was after the renaming. And you are correct a vast majority said do not change the point system, but the ones that wanted the point system changed said that those poles are never accurate anyway.

I do not know that it will ever be explained as to why they decided to reduce the points across the board on bigadv when in reality the points that were being griped about were the values being assigned to the new bigadv WU's. I know they had a meeting about the points but with who I do not know. I would imagine it was the Stanford staff and the DAB which is comprised of a rep. from each of the top 5 teams which are EVGA, [H]ardFolding , Overclock.net, overclockers.com, and Maximum PC Magazine and only 1 or 2 of those really have much of a stake in the bigadv folding. And what they did is not going to stop the problem referred to in Kasson's post of too many people hacking the system and choosing to run WU's that were not designed for them to run. The only thing that is going to stop that is shortening the deadlines or a code that reads CPU signatures.

Anyway as I said earlier it does not really matter to me, the points are just a little added fun and admitley an incentive. But even with the point reduction nothing has really changed bigadv is still making more points than smp and a person with a big bigadv rig is still making more points than a person folding bigadv.

I have been invested quite a lot of money in 7 rigs, 5 i7 9XX and 2 i7-2600k. This is done to obtain a larger production of PPD, and to save power. After downgrading of bonus points do I use approx. 15% more power to achieve the same PPD as before. Previously, I did only GPU folding, but by changing on to the CPU BigAdv folding, a year ago, I lowered my consumption by 50%. A week ago I passed 100,000,000 points. I have been folding since October 2008. Because of the change in bonus points, I now quit folding.


Edit: Correcting an error!

-alias- wrote:I have been invested quite a lot of money in 7 rigs, 6 i7 9XX and 2 i7-2600k. This is done to obtain a larger production of PPD, and to save power. After downgrading of bonus points do I use approx. 15% more power to achieve the same PPD as before. Previously, I did only GPU folding, but by changing on to the CPU BigAdv folding, a year ago, I lowered my consumption by 50%. A week ago I passed 100,000,000 points. I have been folding since October 2008. Because of the change in bonus points, I now quit folding.

Interesting. Why would you invest so much money in points? The scientific value of your work remained the same. Unfortunate to hear, but I wish you the best.

FlipBack wrote:

-alias- wrote:I have been invested quite a lot of money in 7 rigs, 6 i7 9XX and 2 i7-2600k. This is done to obtain a larger production of PPD, and to save power. After downgrading of bonus points do I use approx. 15% more power to achieve the same PPD as before. Previously, I did only GPU folding, but by changing on to the CPU BigAdv folding, a year ago, I lowered my consumption by 50%. A week ago I passed 100,000,000 points. I have been folding since October 2008. Because of the change in bonus points, I now quit folding.

Interesting. Why would you invest so much money in points? The scientific value of your work remained the same. Unfortunate to hear, but I wish you the best.

The only tangible measure of value that donors see is points. It's the only thing that separates a classic WU from a GPU unit or a bigadv unit in the minds of most donors.

Taking points away from all Bigadvs (especially after years of having them in place)...In the 1.8yrs I've been Folding, I don't recall a large ground-swell of people saying they felt that Bigadvs were out of proportion.

No, the -bigadv experiment has not been running for years. Although I thank you for your contribution and respect you as a fellow Folder, 1.8 years of experience does not position you well for a broad perspective.

To those people who just acquired X and Y hardware to implement competitive strategy Z: do you really think this is the first time that the program administrative measures have changed or the first time that a strategy didn't work the way a Folding member thought it would? Feel free to roll your eyes at the following - what part of "beta" did some of you guys not understand, vis a vis big advanced and big-big advanced projects? Sure, I've had my own moments (many of them, ahem ) of bruised ego and dashed schemes for Folding conquest, but I've learned to adjust and move on.

Your X and Y hardware will still be more competitive than the A and B hardware that it replaced. (That includes the 2600K system I just built two weeks ago.)

kirtar wrote:

FlipBack wrote:

-alias- wrote:I have been invested quite a lot of money in 7 rigs, 6 i7 9XX and 2 i7-2600k. This is done to obtain a larger production of PPD, and to save power. After downgrading of bonus points do I use approx. 15% more power to achieve the same PPD as before. Previously, I did only GPU folding, but by changing on to the CPU BigAdv folding, a year ago, I lowered my consumption by 50%. A week ago I passed 100,000,000 points. I have been folding since October 2008. Because of the change in bonus points, I now quit folding.

Interesting. Why would you invest so much money in points? The scientific value of your work remained the same. Unfortunate to hear, but I wish you the best.

The only tangible measure of value that donors see is points. It's the only thing that separates a classic WU from a GPU unit or a bigadv unit in the minds of most donors.

I guess. It just seems odd to quit folding after investing that much money due to a 20% reduction in point numbers. I mean, obviously to invest that sort of money in folding, it's because of the scientific value. Points are just a way to measure contribution and add competition and fun to it. Seems an odd reason to quit to me. But what do I know, I haven't invest hundreds in it yet, just what I already own.

To those people who just acquired X and Y hardware to implement competitive strategy Z: do you really think this is the first time that the program administrative measures have changed or the first time that a strategy didn't work the way a Folding member thought it would? Feel free to roll your eyes at the following - what part of "beta" did some of you guys not understand, vis a vis big advanced and big-big advanced projects? Sure, I've had my own moments (many of them, ahem ) of bruised ego and dashed schemes for Folding conquest, but I've learned to adjust and move on.

War,famine and disease are also all the norm of all human history.

However,the generally accepted strategy is to work to minimize their occurrence, rather than embrace their inevitability.

We've always done it this way (with poor results) is not a compelling argument.

Leonardo wrote:

Taking points away from all Bigadvs (especially after years of having them in place)...In the 1.8yrs I've been Folding, I don't recall a large ground-swell of people saying they felt that Bigadvs were out of proportion.

No, the -bigadv experiment has not been running for years. Although I thank you for your contribution and respect you as a fellow Folder, 1.8 years of experience does not position you well for a broad perspective.

Actually, I wasn't trying to claim a "broad perspective" in my commentary. I was simply saying that in the time that I've been Folding, and having actively participated in my Team's Folding Forum, I have not seen regular complaints about Bigadv points (with exception to the recent 6903 WU). It doesn't take 10 years of experience to be able to take note of that and it appears the Bigadvs have been around at least since November 2009. I don't know when they actually appeared (if it was earlier that that), but that is certainly long enough for people to "feel" like the Bigadvs have been around a long or sufficient time to get used to the points structure. To suddenly change that points structure for all Bigadvs seemingly because a few people had issues with the 6903, just did not seem necessary. That is the gist of what I was saying and is also that of other Folders who have commented about the change.

Without wishing to generalise too much, I think it fair to say that for a lot of folders - particularly those who have joined teams - the team aspect of distributed computing is what attracts them. A team may have its own forum on a site or a sub-forum dedicated to their efforts. There is a herd mentality. Rising to the top of the herd and being in the best herd is part of the attraction.

The way that many DC'ers get their rewards is not through the science - which let's face it is way over most of our heads - but through a Pavlovian reaction to the points system. Points are like the pat on the head that a dog gets.

The internal team competition and the intra-team competition eggs contributors on to go faster, produce more, find the sweet spot in the hardware stakes.

Every time the points have changed somebody gets disappointed. The most recent was the points change when the early FERMI 925/912 units dried up and people who'd rushed out to buy Nvidia 450s suddenly lost half their "production".

Now we have this. The people who had bought i7-920s and more recently 2600K bought them because it was clear there was a huge competitive advantage. Replacing an ageing Athlon X2-3800 with a 2600K running bigadv, enabled my own points to go from 40K ppd overall to nearly 80K ppd. My reward was obvious and led me to encourage my team-mates to do the same.

But the huge power of the 2600K also led me to reconsider the points contributions my other dual-core Intel cpus were making and their SMP clients have now been disabled. They simply weren't worth the electricity or the administration. But that's no different from what happened when I bought dual-core cpus when SMP came out.

This has always been a treadmill and always will. What seems to have happened this time though is that the newest big-bigadv units on the most powerful hardware didn't just offer a big benefit; they appeared to effectively out-produce tens of thousands of older machines doing the oldest units. It was a logarithmic increase.

Some teams which had rigs capable of running big-bigadv had detailed instructions on how to maximise the production of 1m ppd systems. Other teams complained about "obvious cheating" not realising how the points system was rewarding systems with 48 cores. Now because of a handful of massive servers and two specific work units, everyone running ordinary bigadv is having their returns "revalued" as well as the big-bigadv.

Was anyone complaining about a 2600K producing 40K ppd? I certainly didn't see complaints on here about it. Yes, there were complaints that an AMD 1090T could make the deadline but wasn't officially allowed to do so but that wasn't a complaint about the points on offer - quite the reverse.

Teams see the impact on their internal races and on their races with other teams. I have already seen a number of forums claiming this points change is a conspiracy aimed at them specifically.

Changing the points for the 6903/6904 would have been easy, non-controversial and irrelevant to 99.9% of folders. Changing the points for all the bigadv folders has upset a lot of dedicated long-term folders.

HaloJones wrote: But the huge power of the 2600K also led me to reconsider the points contributions my other dual-core Intel cpus were making and their SMP clients have now been disabled. They simply weren't worth the electricity or the administration.

I have a similar attitude. The 2600K made (still makes) so many PPD, I have thought about moving all my C2Qs to a project that more fairly compensates them and simply turning off all my gpus. I long ago turned off all my uniprocessor clients. The problem isn't merely with the 12 core bigadv ppd, it starts at about the i7 with the bigadv WUs after p2684. Perhaps Pande Group has decided that's not an atttitude they want to foster.

GPU's really should not be compared to CPU's totally different type of work and value of work. From what I understand in laymen’s terms, GPU's eliminate paths and find paths to follow. CPU do the intense work of blazing the trail. I could be wrong but that is the way I understand it.