Folding@Home Weekly Team News - 28th October 2005

Soldato
Joined
26 Dec 2002
Posts
9,348
Location
Derbyshire
Folding@Home Team News - w/e 28th October 2005

Current Team Position - 60th
Current Team Score - 7,472,868
Crunched This Week - 235,263



Next on the stomping horizon :cool:

56 Sharky Forums - 27 days
58 Mississippi State University - 22 days
59 kcgeek - 4 days

Teams stomped this week :D

61 www.overclocking.dk
62 Team EDBpriser.dk

Ones to keep an eye on

176 Team Hack-A-Day - 211 days :eek:


This Week's Top Dumpers

Sammael(OcUK) - 80357
DarkendViper - 18927
hornytoe - 9308
_Berserker_ - 8899
sculptor(OcUK) - 8354
KE1HA - 6277
froggy - 6234
Steevo38_(OcUK) - 5959
polandro - 5298
kinnison - 5206
Walter_Anderson - 4531
rich99million - 4304
Sgoaty - 4055
ShaunBrewer - 4043
Ken - 3844
Whitestar (LigerZero) - 3691
Haly - 3192
aXeR - 3190
jamief - 3133
Bigstan - 2748
happytechie - 2550
Trick - 2320
duggyruss - 2289
Mattus - 2282
Staffy - 2139
NeedleGunner - 1953
KillBait - 1922
Kyzia - 1851
PeterNem - 1769
sr4470 - 1527
beatvar - 1353
theredguy - 1351
nicholas_yiu - 1286
dr.who - 1266
MetalGuru - 1082
Dean_Vile - 1059
Wyrdo - 1021
Acolyte - 1005
Pixel - 964


This Week's Top Stompers

cheets64 - 200
seo2india - 93
johannes001 - 86
Zirax - 73
chris.morgan01 - 48
TinkerBell - 25
madpierre - 24
jamief - 23
Sgoaty - 22
mhalsall - 15
polandro - 14
PeterNem - 11
Staffy - 9
Whitewater - 7
KillBait - 5
Pedrinho - 5
Rees - 5
Walter_Anderson - 5
beatvar - 4
Ken - 4
pmbuzz - 4
theredguy - 4
WeeRapscallion - 4
Wyrdo - 4
chronicfathead - 2
Dean_Vile - 2
sr4470 - 2


This Week's Tiffy-Kat Winners

Sammael(OcUK) - 680,000
DarkendViper - 620,000
froggy 260,000
hornytoe - 160,000
Whitestar (LigerZero) - 140,000
happytechie - 120,000
Trick - 60,000
Mattus 50,000
NeedleGunner - 40,000
Walter_Anderson - 30,000
Ken - 30,000
hilly - 30,000
polandro - 20,000
Sgoaty - 10,000
Staffy - 10,000
WeeRapscallion - 10,000
jamief - 5,000
Rees - 2,500


This Week's New Members

cheets64
johannes001
seo2india

A very warm foldy welcome to all of you


Production was a bit lower this week - back to 33K a day average, still great output from all those who haven't got their borgs turned o** for half term and we're still gaining on KCGeek :D
Sorry anyone who's not been mentioned - you know what to do now :)


Stats compiled from Stanford official stats page and EOC stats page
Stomping and Dumping Stats calculated between 16:50 October 21st and 16:40 October 28th
Previous Week's News: October 21st, 14th October, October 7th, September 30th, September 23rd, September 16th, September 9th, September 2nd, August 26th.
 
Last edited:
Ye I need to apologise for my output dropping a little. Unfortunately half-term=Rebuilds so all 450+ machines have been re-installed, meaning those machines out of the 450+ that crunch lost their wu's! PLUS the power was o** for 5 hours today in order for the Electricity board to upgrade our local sub-station (we were drawing too much powahhhhhhh!) :( :( :eek:

Back to full speed now though! :D

Congrats to those with tiffys', welcome to those newbies, and more congrats to those who got a mention (even though I'm a little sour for not getting one myself after taking the team #1 spot! :p )
 
Last edited:
well you managed 80K since last week's news so that's not bad going really, I feel for Whitestar :o :(

edit: oooh ya sneaked in on the sly there Haly - yep a nice bit of team stomping after last week - and with production back up next week hopefully we'll br reeling in some more for the week after too :)
 
I have 2 units that have been going since Tuesday and due to dumpt tomorrow. I think about 1000 points between them so should look better next week. I didn't think there were very large ones about.
 
My first WUs from my newest cruncher should be finishing up this afternoon. It's the dual Xeon beast I described earlier. I haven't fiddled with the fsb quite yet so I'm still running 2x1.6GHz. I've done the proper "wire trick" mods to the cpu sockets but haven't gotten around to the vDIMM yet. I forgot to put -advmethods on the first time I started the services so I've got trash: 2 20xx units and 2 75x units. I'll do better later today. QMDs all around I hope.

We'll see how the system bus likes being hammered by 4 320MB RAM QMD in simultaneity. :D
 
My production is down a bit at the moment due to my laptop being unwell. It will be for some time as well as it needs a new fan which will have to be ordered from the states (can't source one in the UK at all) :(

As I am offshore for another two weeks and then going up to Shetland to see my folks after that, it will be a few weeks before I can get it sorted - if at all.

On a more positive note: As I will be up in Shetland, I will be able to pick up the bits required to get my other P4 back together (since my brother and dad failed miserably). If I have time to get it built before I go back offshore, I should be able to fend off the attack of some of those who are starting to catch up (sorry Haly :p ).

Stan :)

PS: My brother has recently bought a new PC so I will be attempting to borg that as compensation for failing to send the P4 bits down - it won't be on all the time but every little helps :D
 
Last edited:
Bigstan said:
As I am offshore for another two weeks and then going up to Shetland to see my folks after that, it will be a few weeks before I can get it sorted - if at all.
Someone might be willing to help out while you're offshore if you give some more details. :)
 
Thanks again Rich. Damn I want an X2 to get my production up, but I know it's a waste of money as all it'll do is fold..!

BillytheImpaler said:
We'll see how the system bus likes being hammered by 4 320MB RAM QMD in simultaneity. :D

Hehe... Berserker (I think it was Berserker?) did some research into this. He found that running 3 QMDs was faster than 4 on dual Xeons. Your mileage may vary!
 
Berserker said:
I never got around to verifying this, but I suspect the best bet for Prestonias (and indeed any 533MHz system) is one QMD per CPU.

I can conduct a scientific experiment to evaluate whether or not the bus can handle it. I can go a step further and do groups of 3 QMDs and groups of 4 QMDs both before and after the 2x overclock to separately evaluate the effect the double fsb had on the RAM bandwidth.

Shall I? :cool:
 
If you wish. :D

Suggestion: Download four WUs, run them until they converge, then start the testing. 5-10% for each test should be enough (discard the first and last % and average the rest).

It's well known that all but the fastest single CPU HT systems are better off with one QMD at a time. My two fast P4s have the memory bandwidth to cope thanks to tight timings, but most don't.
 
I seem to get the best results (in ppd terms) by running one QMD and one other unit on my HT boxes. It also frees up a bit of memory making the desktop certanly feel more responsive for those 'other' applications that I really should use at work.

HT
 
Back
Top Bottom