11th Annual Folding@Home Holiday Season 3-way Race

[borandi]

While everyone folds, Corsair are doing a giveaway on the main site for the US:

http://www.anandtech.com/show/10859/giveaway-corsair-rm1000i-special-edition-psu

 

[Markfw]

At the rate we are at, we will pass both teams in 1.6 weeks. Too bad the race isn't for 2 weeks.

Unless we can ramp production a lot more !!

 

[bds71]

note: I only signed up with the team yesterday (your total points resets to zero when you switch to a new team). to my knowledge the 24 hr average actually averages over 7 days - since I only have been folding for the team for 1 day my 24 hr avg is WAY low (1/7th) - my actual avg will be closer to 250-300k vs the 38k showing now (38k x 7 = 266k).

how many other new-comers are in the same situation? that could skew the numbers heavily in our favor!

 

[TennesseeTony]

I agree, we are actually gaining on them to the tune of about 15 million per day, times 7 race days, equals 105M, and we are only behind 112M.

We still have the 17 hours left TODAY, pre-race, to make that other 7M points. So I think we will not only beat them for the week, but we shall also overtake them in the very last moments of the race. A DOUBLE Victory!!!

 

[Markfw]

Except I will loose almost 2 million ppd by Friday. I have to return my RMA cards.

 

[TennesseeTony]

Well I'm sure both Teams have last minute joiners, causing these estimates to be skewed, but we seem to have had the edge by a large margin in new blood so far! Here's hoping the last minute crunchers can replace that missing 2M ppd!

 

[bds71]

victory = awesome!

we need 16.03M gain daily to pull ahead of them (starting at the end of today)....currently we are at 10.8M (using 24 hr avg)

tony: where did you get the 15M number from?

 

[TennesseeTony]

As you have said the 24 hour average is based on the last 7 days. I chose to use the last 24 hours actual output: TA-38,703,637 versus Tom's-23,470,734. 15,232,903 difference. The difference (to our favor) was even higher, but Tom's is starting to build a bit of last second steam.

Those stats were 90minutes ago, so by the start of the race (18 more hours of work) we will have produced another 29,027,727 versus Tom's 17,603,050, for a potential (total score) gain of 11,424,677 before the race even starts.

Data for Team Anandtech, data for Tom's.

 

[bds71]

ahhh, I see where you got from now. cool.....looks like we got a real chance of taking them

 

[bds71]

dumb question: what is the actual criteria for winning/losing? how will the winner be determined?

 

[Markfw]

Whoever does the most points in 7 days.

But we want to also pass them on total points, just because.

 

[TennesseeTony]

Points made during the 7 day race with Tom's. Team lifetime totals are not considered, just icing on the cake if we overtake.

Against Hardware Canucks Forum, points made during the entire race (ends Jan 1st). This will include points made while Tom's was with us (against us, actually ).

This is our 11th annual competition, and most of those have simply been an internal race for fun. Once or twice before we competed with another team.

BrainEater came in several months back and challenged us to try and overtake the Canucks, but it was still quite hot out, so we deferred until now, our normal race month.

Then quite by surprise, the charity event against Tom's came about in the last few weeks.

So our 'normal' little friendly internal annual competition has gotten....complicated. And we like that.

 

[bds71]

nice - looks like we got the competition "in the bag" and, as you say, overcoming them in total points in icing on the cake.

 

[Storm Vanguard]

ok guys, every system is go, the rig folded without errors for 12 hours straight, i'm firing up my first 970 for 24/7 operations right now. this should give you around 300,000ppd, guaranteed unless i run into technical problems. also, i'll fire up the second 970 for 5-6 hours per day but no more, unfortunately if i fold at full steam every single fan in my case spins up to maximum speed and i just can't sleep with a vacuum cleaner on beside my bed... there's also a power concern, i'm running all this stuff on a corsair CX600 while waiting for the RMA on my AX860, so i'm a little edgy at the idea that the power supply may give up if i ask him 80-90% load for one month straight. all in all, it will probably be around 400,000ppd from me, i don't think i'll be able to push out much more. still decent though. let's take the crown

 

[Markfw]

BTW, if you want some comedy, read the Toms hardware thread on this. They are a bunch of whiners, and scared they are going to get demolished (which they will)

http://www.tomshardware.com/forum/id-3229962/tom-hardware-team-thread/page-3.html

 

[crashtech]

I used to like Tom's a lot, way back when it was Tom Pabst writing the articles.

 

[indydude345]

I have been experiencing trouble with my computer the past few days, but currently it is up and folding!

 

[Kiska]

Argh... I still don't have thermal paste, nor do I have any tracking on my lga1366 Mobo + cpu for it. So my makeshift pc that will house one of the GPUs isn't starting up until I get thermal paste

 

[ZipSpeed]

Hey guys, all this reading about how this is the biggest Folding competition yet has me wanting to join. So I have decided to join in on the race! So far, I have committed a single GPU (R7 370) and I'm seeing ~84000 PPD. Peter, I'll be folding under ZipSpeed.

It's been awhile since I've done any Folding and looks like the last time was back during the 2013 race. Interestingly enough, I'm still in the top 20 in points for the team (sitting at exactly 20th lol).

As some of you may or may not know, I prefer World Community Grid as my choice for DC projects and I can't guarantee anything, but I may add additional GPUs into the race at a later time.

 

[borandi]

November just finished. 508,534,735 points for the month.

At our current rate of 38m/day, we could get 1,198,119,775 in December (~1.2 billion).

We should make 1 billion a stretch goal. I'll gladly donate to Child's Play / send some of my unused hw to a random participant(s) if we hit that goal.

 

[TennesseeTony]

Welcome back ZipSpeed! I've been jealously eyeballing your scores here and there, wondering if I can ever catch you in a few places (WCG in particular). It's good to 'meet' you!

 

[theAnimal]

Our last 24 hour average is approaching 40M! Good work everyone!

 

[indydude345]

My numbers are promising, 380k PPD with an R9 290. Not bad, from my perspective.

 

[borandi]

I use this as a rough chart of what I should be expecting per card. As long as it's ballpark...

https://docs.google.com/spreadsheets/d/1vcVoSVtamcoGj5sFfvKF_XlvuviWWveJIg_iZ8U2bf0/pub?output=html

 

[alpha754293]

Ok I've dropped back to the 373.06 drivers and will let you know if that fixes the issue with new tasks. Unfortunately it seems to have started a pretty large task estimating 9 hours to completion so it'll take a while to find out.
Changing CPU threads from auto 7 to 6 seems to have got the new task working there too but again I'll have to wait for it to finish before I can tell if it's truly fixed.
Thanks guys!

You can read some of my earlier posts on this thread about how many threads to set your CPU slot to due to issues with Folding@Home's WU assignment server (not being able to assign WUs to slots that only have/use 7, 11, etc. threads).