Folding@Home bigadv EOL 4th Anniversary Challenge: [H] vs TAAT

[[H]Coleslaw]

I DM'd @biodoc if we were still planning on having this event. He said that he didn't have the feeling that TAAT was still interested and for us to see if there was. So with that said, what are your feelings on the event? It would be 1/24-1/31 if it took place.

Edit: I am looking for just TAAT's opinions as [H] was already planning on showing up.

 

[Ken g6]

Primegrid's planning a slightly overlapping, GPU-enabled challenge.

 

[motqalden]

Did we want to change up the dates a bit to avoid any overlap then?
Was really hoping to do this challenge again this year. We can change it up to keep it simple.
No sign ups and just whatever points the team gets over the duration of the challenge

 

[motqalden]

To all you that voted no is there a particular reason for that? Life is too short for people to get upset about DC and I have noticed that a lot of people in the DC community tend to get upset easily for one reason or another. Lets air our grievances and talk it out like reasonable gentleman. There is a lot of history between our teams both good and bad and no reason we can't get along.

 

[Markfw]

To all you that voted no is there a particular reason for that? Life is too short for people to get upset about DC and I have noticed that a lot of people in the DC community tend to get upset easily for one reason or another. Lets air our grievances and talk it out like reasonable gentleman. There is a lot of history between our teams both good and bad and no reason we can't get along.

I do not want to go into particulars. DC are great people, but some of the comments made last year soured us on doing this. I will leave it it that. Those that made the comments know who you are. And competition ? Last year it was not "keep it as close as we can and be fair", but "lets see how we can cheat and win".

I have said too much.

Edit: Hint H's team product was less than 3 billion per month before the race. During it shot to 10 billion. after it went back to 2. Whats wrong with this picture ??

 

[motqalden]

I do not want to go into particulars. DC are great people, but some of the comments hade last year soured us on doing this. I will leave it it that. Those that made the comments know who you are. And competition ? Last year it was not "keep it as close as we can and be fair", but "lets see how we can cheat and win".

I have said too much.

Edit: Hint H's team product was less than 3 billion per month before the race. During it shot to 10 billion. after it went back to 2. Whats wrong with this picture ??

Most of the big points earners for that contest (myself included ) don't regularly fold so the points leading up to the contest is a pointless metric.

About the only thing that you might be able to argue as "cheating" that occurred would have been the fact that we had some folks from another team that decided without consultation to join us for the challenge. We did not actively recruit them or anything they just happened to see the posts about the challenge and joined up.
If you go by the fact that we were by the rules stated we were allowed to have anyone join in at any time and there was no pre sign up for our side. No we didn't tell them they couldn't join us when they offered but how doesn't this fall under lucky recruiting?

EVEN with the fact that those guys joined I was keeping a rough eye on tallies and if we removed the points from those particular individual I believe we would still have come out ahead or it would have been close.


keep it as close as we can and be fair", but "lets see how we can cheat and win".
>>> I actually have a big problem with this. Its a contest so if you change teams to try to make it even then really what is the point. I personally don't mind winning or losing and the big winner a the end of the day is the science.

Going back a few years with this contest you guys have consistently underestimated us based on PPD for the team leading up to the challenge and there has been this push from your team or some members on it to try to add people to our team or make the rules different. For the first contest you gave us less members than you initially wanted to, because we kept saying we were gonna produce more than you figured we would. we surprised you and the team the had a few of your guys on our side ended up winning (I could be wrong on the win part just going from memory)

Last year you still seemed to think that you needed to give us some kind of edge. Frankly speaking it comes across as slightly offensive when you try to give us a major handycap just so that it will be even. I could give a rats tail if i lose or win but a fair race is at least interesting. A contrived race with handicaps or mixing teams to "try to keep it even" accomplishes what exactly?

I will say that my main goal going into last years race was to beat you guys to the point that you would come back the following year and stop with the different rules for thee then for me stuff.
Outside of the 2 or 3 dudes who joined in because they wanted to we really hit it hard recruiting from the forum and got a lot of interest so I guess we pulled more points then you expected but is that really a crime or cheating? If you consider this from the other side of the aisle don't you think that the rule differences can come across as something like a charity handout?

So what I propose here is let skip the recruiting. Both sides brings everything they got without sign ups for the whole duration. Who the heck cares who wins, the science wins and we give more visibility to a project once a year and maybe some people who don't fold will start and maybe old timers that stopped will start again.

 

[motqalden]

And finally to add I kind of fully expect you guys will win if you bring it all but that is OK.

 

[emoga]

If the dates are changed so that they don't conflict with any PG event, then I'd consider changing my vote to yes.

Although, I suppose I could just arrive late to the scene...

 

[motqalden]

If the dates are changed so that they don't conflict with any PG event, then I'd consider changing my vote to yes.

Although, I suppose I could just arrive late to the scene...

I am hoping to focus on PG myself next year so I have no intention of missing half a day or what not of the first challenge. I don't think there is any particular need to not change the dates and we would really only need to shift things forward a day maybe.

 

[crashtech]

To all you that voted no is there a particular reason for that? Life is too short for people to get upset about DC and I have noticed that a lot of people in the DC community tend to get upset easily for one reason or another. Lets air our grievances and talk it out like reasonable gentleman. There is a lot of history between our teams both good and bad and no reason we can't get along.

If you really want to know, my problem is with [H] being a safe haven for anti-Semites. Kyle vacationed me for speaking out against this, and I asked for my account there to be deleted. So now I personally don't want anything to do with [H]. Everyone else can do as they please. I'm not upset particularly, more like disgusted.

 

[motqalden]

If you really want to know, my problem is with [H] being a safe haven for anti-Semites. Kyle vacationed me for speaking out against this, and I asked for my account there to be deleted. So now I personally don't want anything to do with [H]. Everyone else can do as they please. I'm not upset particularly, more like disgusted.

I won't deny there are some bad personalities on the forum but I don't think any of that has ever crossed over to the DC team. Kyle and the forum for the most part have really nothing to do with the DC team at this point and he barely tolerates us tbh. There is a wide mix of personalities in the forum just like in the real world. Hating on our DC team because of some peoples view would be like hating all humans because some of them are anti Semites imho. No bones about it Kyle is a jerk and I generally try to avoid interactions with him. Despite their being some redacted people on [H] it remains a great place to talk about hardware and get help and feedback on things. I think most of the problem people tend to hang out in the restricted areas of [H] where all the political threads are.

Love it or hate it [H] will pretty much tolerate any behaviour outside of blatantly illegal so you can argue its a safe haven for lots of people with a wide variety of opinions.

I appreciate your feedback on this and this hadn't even been something that I would have considered being a factor in someones decisions regarding these topics.



No profanity allowed in the tech forum.


esquared
Anandtech Forum Director

 

[[H]Coleslaw]

The date isnt set in stone. It was assumed based on a previous post by biodoc. We can adjust it however needed.

Also keep in mind that the above screenshot from PrimeGrid is from their discussion thread and not yet posted as complete in its normal spot. There is still one more to be determined and thus more edits may occur as they should. Just an FYI for those planning to partake in the series.

And my vote on the poll wont count. I just needed to vote to see the results.

 

[biodoc]

I voted yes as long as we keep it simple and run it as all of TAAT vs all of [H]. We should also adjust the start date to avoid conflict with the PG challenge.

 

[Pokey]

Last year was more like an ambush than a race. My private comment last year: "I won't do that again."
So I'll pass.

 

[pututu]

I voted yes as long as we keep it simple and run it as all of TAAT vs all of [H]. We should also adjust the start date to avoid conflict with the PG challenge.

Agree we should keep this a simple race between TAAT team and [H] team, no individual signup, just team stat. Perhaps Jan 25 12:00 UTC to Feb 1 12:00 UTC would be a compromise? It's less than 12 hours overlap between PG CNY challenge and Feb PG TDP (28 days of marathon) challenge. TAAT has a head start with the momentum carrying over from the FAH New Year Race, so I think the extra one day of folding won't hurt much.

Last year was more like an ambush than a race. My private comment last year: "I won't do that again."
So I'll pass.

Edit: looks like biodoc proposed ending the TAAT race on Jan 21st.
If you are in the TAAT FAH New Year race starting tomorrow, you are indirectly participating in the BigAdv race, well at least for 6 days. . In last year annual FAH holiday race, you guys were pumping up production during the last few days of the race. If similar pattern holds true, it's going to be tough for us to catch up .

To all TAAT and [H] participants: At the end of the day, it is the science that benefits. We still have a long way to go in the cancer cure and degenerative diseases research, so look at this BigAdv race as supporting the good cause. Let's end this insignificant squabble or misunderstanding and focus on supporting FAH via attracting more people to join in on both sides of the aisle. One way to do this is to have team competitions while having fun. On [H] side (no secret) we will continue to post messages early on in our [H] forum including [H] GenMay forum. The FAH team competition is what attracts people in supporting the good cause.

BTW, this race is called BigAdv EOL race because a long long time ago, [H] was a powerhouse in cpu folding. Stanford ended BigAdv and focus more on gpu folding and that pissed off a lot of [H] folks and most quitted. Yeah, we were a big team back then.....

 

[Endgame124]

That F@H chart video would have been very interesting if it didn’t stop at 4/1/2020

 

[Assimilator1]

That F@H chart video would have been more interesting if it didn’t stop at 4/1/2020

Corrected , but largely agreed, I was looking forwards to the big uptick for spring 2020, alas it wasn't to be.

 

[Skillz]

I posted a comment asking for a new video that shows data up to the current date.

If more people ask for it, they might make it happen.

 

[[H]Coleslaw]

Right now it is a tie. We need some more votes. Just to reiterate, I am only counting TAAT votes as this is a team decision.

 

[crashtech]

I have changed my vote to abstention. I am not going to participate, though.

 

[motqalden]

I'm really sad to hear that some of you don't want to move past this. If it helps I would like to apologise for anything that I may have done or said with the last years contest that upset someone. Can we not shake hand and move forward here?

 

[[H]Coleslaw]

I meant to post earlier but my primary box took a dump and spent the whole day trying to figure out the issue. So, if @crashtech is changing the vote to abstain then the votes are in favor of having the event. The dates will be changed to not conflict with PG challenge but we are not skipping February's event since it is the whole month. I figured bumping it back to the 26th would give enough head room for people to finish up PG and then some time to switch things over and get stable. Please let me know what you guys would prefer and we can work it out as [H] as a team isn't that hung up on the PG challenges this year.

 

[Assimilator1]

26th is fine by me .
Do we have any idea on numbers of participants of the teams?

but a fair race is at least interesting.

Everyone wants a fair race, but the trouble here is it means different things to different people, fair in your case seems to mean unmodified/natural team membership/output, let the race happen as is. Where as fair for many people here means a close race, regardless of who wins, so swapping members around etc is quite acceptable.
I lean towards a close race idea, as a close race is interesting, but unless by sheer chance membership+output is similar between teams, it's exceedingly unlikely I think!
I don't think where one team is obviously going to trounce another, that is going to be much fun.
As far as I can see the 2 ideas for a race are mutually exclusive, but I'm up for it this year!

The other trouble it seems is estimating ppd (as a recent tense discussion in the F@H race thread highlighted), whether due to variance in what rigs the big guns can run, combined with ppd variance between F@H projects, it makes it very hard (impossible?) to get even a decently close estimate.

But were team/ppd modification be considered necessary, it could be achieved using a multiplier to one team's output rather than swapping members. After all, I think we all want to be in our own team?

 

[[H]Coleslaw]

26th is fine by me .
Do we have any idea on numbers of participants of the teams?


Everyone wants a fair race, but the trouble here is it means different things to different people, fair in your case seems to mean unmodified/natural team membership/output, let the race happen as is. Where as fair for many people here means a close race, regardless of who wins, so swapping members around etc is quite acceptable.
I lean towards a close race idea, as a close race is interesting, but unless by sheer chance membership+output is similar between teams, it's exceedingly unlikely I think!
I don't think where one team is obviously going to trounce another, that is going to be much fun.
As far as I can see the 2 ideas for a race are mutually exclusive, but I'm up for it this year!

The other trouble it seems is estimating ppd (as a recent tense discussion in the F@H race thread highlighted), whether due to variance in what rigs the big guns can run, combined with ppd variance between F@H projects, it makes it very hard (impossible?) to get even a decently close estimate.

But were team/ppd modification be considered necessary, it could be achieved using a multiplier to one team's output rather than swapping members. After all, I think we all want to be in our own team?

That is some good insight. The problem I think we have on our end is most don't run FAH anymore or ever. So, getting a "fair" estimate probably won't really happen without just guessing based off of the database itself. Then guessing again based on how you want to make your setup efficient. For example, I typically run mine at 60% power. That is where I intend on running them but if we are getting left in the dust, I may just say screw it and let them run all out. So, there can be a large difference. Then you have the folks that want to pay for cloud servers and don't care how much it costs as long as they win. I don't care for this method but I don't condemn people for doing it. I see no difference in spending hundreds of dollars on cloud servers than spending hundreds of dollars on hardware and electricity. Either way, it is the ones with the biggest wallets that win.

I do agree a close competition is more fun. However, there just isn't any good way to satisfy everyone in this realm. I also like to see people bring everything they got rather than pulling punches. It isn't any fun for me if someone gives me an illusion of winning. Personally, I would rather be left in the dust than to be given false hope. (I'm used to be left in the dust as I typically have old hardware.)

Keep the ideas and suggestions coming. I definitely like getting the voice of the people when it comes to these things.

 

[[H]Coleslaw]

Since we didn't get much discussion on the topic, we are going to make it January 26th - February 2nd.