Folding@Home Weekly Team News - 13th December 2007

Team OcUK Folding@Home News 6th - 13th December 2007

Current Team Position: 25th
Current Team Score: 75,423,611
Crunched This Week: 1,168,108 (24hr avg. 166,873)
Change from Last Week: -80,083


Next on the stomping horizon

21. Dutch Power Cows - 1,217 days (+575 days)
18. DSL Reports Team Helix - 1,069 days (+341 days)
19. ABXZone.com - 656 days (-85 days)
20. p2p-community - 493 days (+60 days)
24. Overclockers Club - 10 days (-3 days)

Teams stomped this week

none this week

Ones to keep an eye on

27. Overclock.net - 22 days (-25 days)
30. 2ch@PS3 - 142 days (+116 days)
44. PlayStation.com Forums Team - 761 days (-431 days)

We are currently at 23rd in team production, down one place from last week


This Week's Top Crunchers

1 . magman - 115,443
2 . KE1HA - 82,885
3 . Cob - 72,927
4 . Whitestar (LigerZero) - 70,665
5 . snapsh0t - 62,105
6 . Biffa - 59,456
7 . ms9cw - 43,793
8 . lemonman - 39,931
9 . 2bullish - 39,131
10 . Babyface_UK - 34,457
11 . Bigstan - 33,271
12 . Neil - 32,616
13 . jaric - 32,315
14 . Gazzza - 28,960
15 . DreederOcUK - 25,884
16 . SiriusB_[OcUK] - 25,013
17 . dan1987 - 19,360
18 . solarisfireball - 17,703
19 . bossk128 - 17,600
20 . wints - 17,600
21 . Unforgiven - 16,830
22 . polandro - 15,439
23 . Borged_by_MGP - 12,320
24 . penfold - 12,292
25 . Mattus - 10,766
26 . hornytoe - 10,560

27 . totalise3 - 8,495
28 . Huddy - 8,228
29 . Rincin - 7,478
30 . MattBrown - 6,638
31 . rich99million - 6,052
32 . Poolybit - 5,660
33 . cokecan72 - 5,622
34 . weescott - 5,557
35 . lightning512 - 5,202
36 . verbal - 4,793
37 . Buxx - 4,502
38 . FatRakoon - 4,304
39 . Nationwide_Access - 4,240
40 . vent - 4,212
41 . ken1307 - 4,043
42 . lopkinfop(OcUK) - 3,630
43 . blitz2163 - 3,539
44 . [SIR]_Petulant - 3,520
45 . _Viking_ - 3,520
46 . marktime - 3,520
47 . sculptor(OcUK) - 3,520
48 . Silent_Bob - 3,520
49 . oceaness - 3,329
50 . BillytheImpaler(OcUK) - 3,029

51 . growse - 2,901
52 . DeQuavus - 2,828
53 . Pilgrim57 - 2,760
54 . Slinwagh - 2,679
55 . biteme99 - 2,570
56 . Trazzall - 2,500
57 . davejuk - 2,478
58 . jeremysmith123 - 2,362
59 . Saetia - 2,350
60 . NOIR - 2,223
61 . steve258 - 2,170
62 . FTW - 2,144
63 . pikeypriest - 2,100
64 . froggy - 2,008
65 . Gunda - 2,006
66 . feriso - 1,900
67 . 355F1512TR - 1,899
68 . br - 1,892
69 . hikaru452 - 1,881
70 . Frogskin - 1,880
71 . nikoPS3 - 1,863
72 . KALIBR(OcUK) - 1,816
73 . schumi84 - 1,760
74 . Steevo38_(OcUK) - 1,760
75 . Freefaller - 1,706
76 . mikeymike - 1,681
77 . _Berserker_ - 1,620
78 . RD - 1,557
79 . diogenese - 1,523
80 . deej - 1,456
81 . casale - 1,393
82 . mish86 - 1,385
83 . Cathode - 1,338
84 . piggott - 1,317
85 . Crowze - 1,292
86 . A_N_Other - 1,255
87 . Ever - 1,217
88 . mark001282 - 1,171
89 . Zymurgy - 1,164
90 . mB0rg - 1,156
91 . Tumbler^ - 1,142
92 . sammy7 - 1,136
93 . JackWolk - 1,129

94 . atsushi - 935
95 . Carnage - 909
96 . ac90 - 845
97 . Pedro_Legend - 838
98 . kinnison - 763
99 . teo.e - 754
100 . ThyFlame - 695
101 . dalin80 - 689
102 . Jagman2k - 686
103 . KillBoY_UK - 686
104 . Montiyouk - 682
105 . Jonathan_CTI_Amesbury - 665
106 . Brown_Sparrow - 618
107 . Garp - 584
108 . zenkohdo - 562
109 . niko - 555
110 . SquAmy - 529
111 . MajorPart - 514
112 . Adnams_Drinker - 500
113 . Dean_Vile - 500
114 . eddie - 500
115 . Galifrey - 500
116 . LiquidFox - 500
117 . TheCrow - 500
118 . jidh007 - 424
119 . Joe42 - 412
120 . FEICHAN - 401
121 . BenjiSayers - 394
122 . theheyes - 394
123 . hawk - 383
124 . PiKe - 372
125 . Kenstheman - 370
126 . Assasin-uk - 343
127 . PS3zappatore - 340
128 . tony - 340
129 . ShaunBrewer - 338
130 . MaxenceLurin - 300
131 . moreno10 - 300
132 . [madoldtory] - 292
133 . Leon_Webster - 292
134 . aceface57 - 265
135 . Scratchy - 250
136 . Northwind - 246
137 . Snapshot - 220
138 . fluiduk - 218
139 . MadFruit 218
140 . Psycho_Sonny - 206
141 . WoZZeR - 206
142 . Limehaus - 188
143 . Cameron - 186
144 . Paul_Holliday - 186
145 . takku - 186
146 . xyphic - 186
147 . happytechie - 168
148 . loiic_corporation_ - 132


Single Machine Folding League

Colour Key: Single Core, Dual Core, Quad Core, PS3, Unclarified

1 . jaric - 32,315
2 . SiriusB_[OcUK] - 25,013
3 . dan1987 - 19,360
4 . Unforgiven - 16,830
5 . Borged_by_MGP - 12,320
6 . Rincin - 7,478
7 . cokecan72 - 5,622
8 . _Viking_ - 3,520
9 . marktime - 3,520
10 . Silent_Bob - 3,520
11 . Pilgrim57 - 2,760
12 . Slinwagh - 2,679
13 . Frogskin - 1,880
14 . schumi84 - 1,760
15 . diogenese - 1,523
16 . Cathode - 1,338
17 . Zymurgy - 1,164
18 . Carnage - 909
19 . Jagman2k - 686
20 . Adnams_Drinker - 500
21 . LiquidFox - 500
22 . BenjiSayers - 394
23 . Limehaus - 188

Cob had just a little help this week.... again

This Week's Top Stompers

atsushi - 508
KillBoY_UK - 468
eddie - 413
hawk - 380
PS3zappatore - 352
MaxenceLurin - 342
Cathode - 129
teo.e - 127
moreno10 - 119
Crowze - 81
Slinwagh - 65
tony - 50
lopkinfop(OcUK) - 35
fluiduk - 31
weescott - 29
loiic_corporation_ - 28
niko - 28
jeremysmith123 - 22
Nationwide_Access - 21
Pedro_Legend - 19
PiKe - 19
nikoPS3 - 17
Silent_Bob - 17
ac90 - 16
mB0rg - 16
schumi84 - 16
JackWolk - 13
steve258 - 13
NOIR - 12
355F1512TR - 11
mish86 - 11
Northwind - 10
takku - 10
wints - 10
zenkohdo - 7
DeQuavus - 6
FEICHAN - 6
Frogskin - 6
LiquidFox - 6
SquAmy - 6
xyphic - 6
hikaru452 - 5
Adnams_Drinker - 4
Cameron - 4
dan1987 - 4
KALIBR(OcUK) - 4
pikeypriest - 4
sammy7 - 4
solarisfireball - 4
biteme99 - 3
Poolybit - 3
bossk128 - 2
Ever - 2
FTW - 2
jidh007 - 2
Montiyouk - 2
ms9cw - 2
TheCrow - 2
Zymurgy - 2
Borged_by_MGP - 1
Brown_Sparrow - 1
Carnage - 1
casale - 1
Garp - 1
ken1307 - 1
Limehaus - 1
mark001282 - 1
MattBrown - 1
penfold - 1
Saetia - 1
Scratchy - 1
SiriusB_[OcUK] - 1
snapsh0t - 1
ThyFlame - 1
totalise3 - 1
Whitestar (LigerZero) - 1
WoZZeR - 1


This Week's Tiffy-Kat Winners

Whitestar (LigerZero) - 5,000,000
magman - 2,400,000
KE1HA - 1,900,000
Bigstan - 1,700,000
ms9cw - 1,600,000
Cob - 1,100,000
lemonman - 1,000,000
jaric - 850,000
SiriusB_[OcUK] - 800,000
Biffa - 750,000
BillytheImpaler(OcUK) - 425,000
feriso - 375,000
Mattus - 350,000
Rincin - 325,000
marktime - 225,000
mikeymike - 200,000
bossk128 - 175,000
cokecan72 - 175,000
solarisfireball - 150,000
dan1987 - 100,000
wints - 80,000
Poolybit - 70,000
Carnage - 50,000
dalin80 - 50,000
Silent_Bob - 25,000
weescott - 25,000
steve258 - 20,000
JackWolk - 15,000
jeremysmith123 - 15,000
lopkinfop(OcUK) - 15,000
TheCrow - 15,000
mB0rg - 10,000
SquAmy - 10,000
ac90 - 7,500
FEICHAN - 7,500
Slinwagh - 5,000
teo.e - 1,000
tony - 1,000

Congrats to all the Tiffy-Kat winners


This Week's New Members

atsushi
budge1972
daniel
eddie
hawk
KillBoY_UK
macayo
MaxenceLurin
weeedy

A great big warm welcome to the team!

If you haven't done so already then why not check out the Folding@Home Setup Guide and the Quickie Guide for more foldy-tastic tips especially for when you're in a rush


We're missing some big production at the top, still all in all some blinding production as a team - way to go all
Since the main stats page was down for maintenance and EOC had their big hiccup around this time last week I wouldn't worry overly about all the red, not unless we get the same next week anyway
If you've not been mentioned either I've made a mistake or your stats haven't changed since last week - give a shout and let us know you're still alive
***Only got one machine crunching? - Have a looky here***

Stats compiled from Stanford official stats page and EOC stats page - also helped out this week by Kakao Stats
Stomping and Dumping Stats calculated between 19:30 December 6th and 19:40 December 13th

Previous Weeks' News:
December 6th, November 29th, November 22nd, November 15th, November 8th, November 1st, October 25th, October 19th, October 11th, October 4th, September 27th, September 20th, September 14th, September 7th, August 31st, August 24th, August 17th, August 10th, August 3rd, July 26th, July 19th, July 12th, July 5th, June 28th, June 21st, June 14th, June 7th, May 31st, May 24th, May 17th, May 10th, May 3rd, April 26th, April 19th, April 12th, April 5th, March 29th, March 15th, March 8th, March 1st, February 22nd, February 15th, February 8th, February 1st, January 25th, January 18th, January 11th, January 4th

2006 News Archives:
December 28th, December 21st, December 14th, December 7th, November 30th, November 23rd, November 16th, November 9th, November 2nd, October 26th, October 19th, October 12th, October 5th, September 28th, September 21st, September 14th, September 7th, August 31st, August 24th, August 17th, August 10th, August 3rd, July 27th, July 20th, July 13th, July 6th, June 29th, June 22nd, June 15th, June 8th, June 2nd, May 26th, May 19th, May 12th, May 5th, April 28th, April 21st, April 14th, April 7th, March 31st, March 24th, March 17th, March 10th, March 3rd, February 24th, February 17th, February 10th, February 3rd, January 27th, January 20th, January 13th, January 6th

Single Machine Crunchers:
Adnams_Drinker, Andrew, BenjiSayers, BigDannyO, booyaka, Borged_by_MGP, Carnage, Cathode, Cob, cokecan72, dan1987, diogenese, Emily, Frogskin, gt_junkie, Jagman2k, jaric, Limehaus, LiquidFox, marktime, Pilgrim57, Rincin, Silent_Bob, SiriusB_[OcUK], Slinwagh, Unforgiven, Viking, Yeggstry, Zefan, Zymurgy ...

I've made a few changes to the tiffy bands - hoping to revalue them a bit as they were starting to be as easy to get as an arts degree

The new bands I've decided on are as follows:

My first tiffy = 1k
2,500+ = every 2.5k
15,000+ = every 5k
30,000+ = every 10k
100,000+ = every 25k
500,000+ = every 50k
1,000,000+ = every 100k
2,500,000+ = every 250k
5,000,000+ = every 500k


I think that's a step in the right direction, I was planning on waiting until Stanford did their long awaited points rejig, if it changes things in a big way we can rethink them again then

As always I'm open to ideas if anyone has any views on the tiffy situation or anything else in the news - just give a yell


edit: This is what the tiffy's used to go like:

My first tiffy = 1k
2,500+ = every 2.5k
15,000+ = every 5k
30,000+ = every 10k
100,000+ = every 20k
200,000+ = every 25k
1,000,000+ = every 50k

Points rejig?

Noooooooooooooooooooooooooooooooooooooooooooooooo!

1 . magman - 115,443
2 . KE1HA - 82,885
3 . Cob - 72,927

Click to expand...

Berserker-powar at full strength!


Ta Rich!

8 . lemonman - 39,931 and to think i was a few k off 30 last week

Thanks for the news rich, and a big welcome to all you new folks - any questions just ask.

Edit: oh and Parp to the post Berserker-powar Cob, your not geting away from me this time

Cob said:

Points rejig?

Noooooooooooooooooooooooooooooooooooooooooooooooo!

Click to expand...

If it ever happens, more credit to the GPU folk would be nice, if there are actually any left

Hmm, not an amazing week from me. Lost a WU yesterday. Was only 20% in, but still.

Haven't posted my news today. I have the results [if a little later than I wanted] but no display yet. The php scripts are under renovation. Think of a car in bits and you should get the idea

Cleaning up the scripts to make it a bit more efficient and also adding a bit more functionality to more closely resemble the current news. I have also bookmarked this page, so I remember to tweak the tiffy bands in my code.

Thanks for the news rich as always and welcome to the new blood. *puts in teeth*

We touched on this a wee bit above, but I think it might be interesting to go into a little greater depth. They Stanford boys have been hinting at a revamp of the points system for quite some time on account of the outdatedness of the single core benchmark machine and the rise of high performance clients on GPUs and the PS3 and in the future on Clearspeed.

That makes me think about how they're going to structure the system to

• Be fair to contributors, however they choose to perceive that
• Be respectful of the existent points of current contributors, e.g. don't suddenly make the benchmark score 10k PPD, thus making past contributors think their effort was chopped liver
• Add incentives for people to use machines and clients that give the best scientific performance, not just points

That last one is particularly relevant to me. I had a go with the GPU client which is about 10x as productive as a standard client. I usually run the Linux MSP client on the machine. We'll say that that gives 4x the performance of a standard client to be generous. The GPU WUs give me less PPD than the SMP WUs. If the GPU WUs scored better they'd be getting more than twice the productivity out of me in terms of scientific output.

How would you enterprising chaps address the problem?

Thanks Rich Not a bad week for me considering that I've been exercising my new 8800GT quite a bit If only I could fold on it...

Anyway... the way I understand it, a GPU can do 10x as many calculations as a standard client, but the range of calculations the GPU can do is much more limited. So I think it's reasonable for GPU points to sit in line with CPU output, if you rate processors on a 'utility' basis rather than pure number-crunching speed. It's not fair to give GPUs loads of points when the work the CPU is doing couldn't even be done on the GPU. Move GPUs too far ahead and the hardcore folders will build rigs around Athlon 1100s and stuff them with X1900XTs... when that happens the project has problems!

IMO the SMP client points are too high. There does need to be some incentive to run SMP rather than 2 or 4 standard clients, but when my E6300 scores 1900ppd on SMP and 400-800ppd on normal clients, the difference is too great.

Change the benchmark machine to a C2D - the current P4 2.8 is old hat. Benchmark both SMP and standard units on that system and then give SMP units a 50% bonus. Give GPU units a 50% bonus too, ensuring that a mid-range GPU scores similar points to a mid-range CPU on SMP. Then normalise those points so that a reasonable CPU gets 110ppd like it used to, bringing SMP and GPU points in line with those for previous bonus WUs.

rich99million said:

11 . Bigstan - 33,271

Click to expand...

That's better

Should break into the top 10 dumpers next week (assuming all is well chez Stan).
As per my thread, getting back up to a reasonable output again. When I get back home in a month, I should be able to add the E6300 to the farm and get a better board for the E6600 so I can oc it a bit - should hopefully get me close to the 10k ppd mark and keep those pesky kids off my tail

Nice effort from the team

Thanks as always to Rich for the news and a warm foldy type welcome to the new folders

FOLD ON TEAM 10

Mattus said:

when that happens the project has problems

Click to expand...

I'll play devil's advocate...

Why would that indicate a problem? Shouldn't points be awarded on the basis of the amount I have contributed to the study at hand? If such a mishmash rig can put out 40x the amount of useful results as a C2D in the same period of time should not it be awarded 40x the points?

This is the crux of the issue I mentioned above. Stanford would benefit from me crunching on my GPU. I would benefit from crunching on my CPU. Stanford is losing here.

BillytheImpaler said:

Why would that indicate a problem? Shouldn't points be awarded on the basis of the amount I have contributed to the study at hand? If such a mishmash rig can put out 40x the amount of useful results as a C2D in the same period of time should not it be awarded 40x the points?

Click to expand...

I see what you're saying - but what I meant by 'problems' in this case was that if GPUs got 40x the points of CPUs, every serious folder would run exclusively GPU clients (bearing in mind that each GPU client hogs a CPU as well.) Then Stanford would struggle to do the science which can only be done on a CPU because there'd be no CPU folders left! AFAIK the balance of work available is well in favour of CPUs. Stanford would benefit from you running your GPU, but if everybody ran their GPUs instead of CPUs then a lot of the work wouldn't get done.

That was why I suggested having a mid-range GPU and a mid-range CPU giving roughly the same points. Some people will have a higher-end CPU than GPU and will thus run the CPU client, and vice versa. Hopefully then we could come to some form of balance in available CPUs and GPUs.

Mattus said:

I see what you're saying - but what I meant by 'problems' in this case was that if GPUs got 40x the points of CPUs, every serious folder would run exclusively GPU clients (bearing in mind that each GPU client hogs a CPU as well.) Then Stanford would struggle to do the science which can only be done on a CPU because there'd be no CPU folders left! AFAIK the balance of work available is well in favour of CPUs. Stanford would benefit from you running your GPU, but if everybody ran their GPUs instead of CPUs then a lot of the work wouldn't get done.

Click to expand...

I think they'd have enough CPU clients left to get the job done. I don't think the points should be 40x more or whatever the figure is but it should be brought more into line with the SMP output (which shouldn't have been made so high in the first place) to reflect the extra cost of buying and running a GPU crunching system.

To add to the debate - as soon as smp goes live (instead of open beta) the deadline will be increased and the bonus points will be lost anyway. Thereby ending the inflated SMP credit (not that I'm complaining)

What every the out come I'll still be doing it, as afterall the science is the 'why' I fold, the points is just a bonus that strengthen the folding community.

The SMP client doesn't get nearly enough points!

Only kidding

As I am given to understand, and as shadow has pointed out, SMP enjoys lots of points because of its BETA status. The only problem is, although in theory the points will go down when it leaves BETA, how likely is that? I have not heard one sniff of it leaving BETA. Even the GPU client is still BETA the last time I checked.

I agree GPU pointage should go up. Given the speciality of the calculations it can do and how much it costs in electricity, noise and heat to run it, GPU folders deserved more. I don't think they should be getting 40x the points, as overnight everyone would be GPU folding and getting 40K PPD! With points like that no one would touch a CPU again.

I think the biggest change needs to be the standard client. I mean come on, you are lucky to reach 400PPD on a standard client. They should increase the points there a little.

Whatever Stanford decides, they need to make sure it is well balanced changes. The last thing the community needs is everyone getting shafted. If points sky-rocket it will make the last 800,000 points seem wasted effort. And if they make the new points rock-bottom you're looking at people on big points being almost permanently ahead as the lower-end crunchers could never hope to catch up.

To be honest though they don't need a massive overhaul. Just increase the points for GPU and Standard and everyone will be happy.

SiriusB said:

To be honest though they don't need a massive overhaul. Just increase the points for GPU and Standard and everyone will be happy.

Click to expand...

I think that this is likely what will happen. Perhaps the PS3 points will go up as well since PS3 participation has fallen a bit.

That's it from me for a while, guys. I'm off for a week skiing in Aspen Colorado with my sister's ski club. Wish me luck.

BillytheImpaler said:

That's it from me for a while, guys. I'm off for a week skiing in Aspen Colorado with my sister's ski club. Wish me luck.

Click to expand...

Have fun

Wouldn't it be better if SMP/GPU points went down rather than standard points going up? If you could suddenly get 500ppd on anything half-decent then it'd devalue people's previous points, especially if they'd been running the standard client.

Mattus said:

Have fun

Wouldn't it be better if SMP/GPU points went down rather than standard points going up? If you could suddenly get 500ppd on anything half-decent then it'd devalue people's previous points, especially if they'd been running the standard client.

Click to expand...

Watch you don't do too much impaling, Billy


Whilst it makes more sense to back off the SMP points a little I expect the solution will be to increase the others so they're more in line - after all how many currently active crunchers will they lose from boosting points compared to how many may get upset and leave if their expensive new rig suddenly gets its PPD cut down?

Guess we'll have to wait and see and adjust if we have/want to

If Stanford adjusts points in such a way we don't agree with, I will create a new stats site that adds a few K PPD to everyone's stats

rich99million said:

Whilst it makes more sense to back off the SMP points a little I expect the solution will be to increase the others so they're more in line - after all how many currently active crunchers will they lose from boosting points compared to how many may get upset and leave if their expensive new rig suddenly gets its PPD cut down?

Click to expand...

Pretty much what I was thinking Rich. I'd be most upset if I'd spent over 2 grand replacing my A64 farm with Core 2s only to have the advantages I paid for removed. I agree that there is a serious inequality at the moment but I'm trusting Stanford not to be silly about it. I'm sure I'm not the only one who upgraded the farm specifically to boost the amount of work I could do and subsequently, my score.