DF Daily (01/28/03)

[m0ti]

Hello Everyone,

Well the protein change is just getting started. The DF web-site should be back up soon. Make sure you've got the updated client! BTW, my computer seems to have stopped folding as it tries to get the last update and failed. Bunker pointed out a solution (disable net-access for DF). Other possible solution is to use dfDetect (otherwise DF exits out since the server tends to be unavailable the first few hours after a protein change) to keep the client running and grab the update whenever possible, however the client won't be folding during this time. This is only a problem when starting up the client when it tries to get in touch with the server when it starts up, so if your client hasn't been restarted you've got nothing to worry about (mine was... restarted the computer).

Looks like the protein is 96AA's... it should probably be faster as well.

ATWE

Not in yet.

Milestone Makers

ATWE Robor hits 200M folds! :Q:Q
TA Rescue Team hits 20M folds!

AppleTalking hits 4M folds!
grunter hits 2M folds!
Deeko hits 500K folds!
levassk hits 400K folds!
None (one of 3 None's on the team) hits 100K!

Congrats to the milestone makers!

FoBoT: A little bit of a drop in production today... perhaps due to the protein changeover? Anyhoo, 8.4M is still good for 5th, which ain't too bad at all. Looks like you finished off the last protein with the 3rd best fold! WTG!

non-FoBoT:
Top Folder: ATWE Robor with 200.8M folds and 59th!
Best RMS: ATWE Terrahawk finishes off the previous protein with the 7th best fold! Big congrats, man!
Best Production: ATWE Robor with 4.8M folds and 15th overall!

Producers Of the Day

FoBoT 8.4M
ATWE Robor 4.8M
ATWE Terrahawk 3.6M
ATWE LastKnight 1.9M
ATWE bfour 1.6M
ATWE Insidious 1.5M
ATWE ReaganCRW 1.4M
TA Rescue Team 1.2M
ATWE cbcrus 1.1M
LANMAN 1.0M

Congrats to the milestone makers!

TeAm Stuff

Overall, we've had some top notch production over this last protein and got a couple of new members, and truly a great effort made by one and all. Let's keep up the good work for the next protein!

Daily Production: 33.7M folds and 4th overall. Rechenkraft.net and FreeDC have both stepped it up. They're at 45.4M and 46.3M respectively. We've definitely got our work cut out for us!

See everyone tomorrow!

Good luck to everyone, and I hope you all have a smooth transition to the new protein!

 

[FoBoT]

Originally posted by: m0ti

Looks like the protein is 96AA's... it should probably be faster as well.

i don't think so, on the last protein my laptop did right at 2 per second, on this one, it is only doing ~1.5

so i think we will see about 25% less production/#'s

 

[Terrahawk]

Good stuff TeAm! I think that thanks to this change, I am losing a whole lot of CPUs - duallies still have problems :frown:

Oh yeah... and check this out



12:52pm up 49 days, 1:50, 3 users, load average: 1.12, 1.06, 1.16
89 processes: 85 sleeping, 2 running, 0 zombie, 2 stopped
CPU0 states: 47.0% user, 2.3% system, 47.4% nice, 49.1% idle
CPU1 states: 51.0% user, 0.0% system, 51.3% nice, 48.1% idle
Mem: 1545244K av, 1446448K used, 98796K free, 0K shrd, 592424K buff
Swap: 2097136K av, 2336K used, 2094800K free 478896K cached

PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND
31337 gkhoo 0 19 87996 85M 1912 R N 99.9 5.6 22:42 foldtrajlite
31395 gkhoo 20 0 992 992 768 R 2.3 0.0 0:01 top
1 root 20 0 92 88 60 S 0.0 0.0 0:34 init
2 root 20 0 0 0 0 SW 0.0 0.0 0:00 keventd
3 root 20 19 0 0 0 SWN 0.0 0.0 0:02 ksoftirqd_CPU0
4 root 20 19 0 0 0 SWN 0.0 0.0 0:02 ksoftirqd_CPU1
5 root 20 0 0 0 0 SW 0.0 0.0 0:07 kswapd
6 root 3 0 0 0 0 SW 0.0 0.0 0:00 bdflush
7 root 20 0 0 0 0 SW 0.0 0.0 14:49 kupdated
8 root 20 0 0 0 0 SW 0.0 0.0 0:00 kinoded
9 root 0 -20 0 0 0 SW< 0.0 0.0 0:00 mdrecoveryd
13 root 20 0 0 0 0 SW 0.0 0.0 0:08 kreiserfsd
79 root 20 0 0 0 0 SW 0.0 0.0 0:00 kjournald
395 root 20 0 600 596 488 S 0.0 0.0 0:09 syslogd
398 root 20 0 1172 1168 372 S 0.0 0.0 0:00 klogd
466 bin 20 0 420 412 328 S 0.0 0.0 0:00 portmap



Process ID for foldtrajlite look familiar?

 

[Insidious]

Thanks m0ti!

Sorry for the late ATWE scores tonight everyone.... was busy getting myself updated and making the new spreadsheet for Best RMS tracking. (m0ti)

Here are the ATWE scores:

as of: Tue Jan 28 19:10:49 2003 EST
Tickets suspended until 1-29-03 update (begin watching then)

Position * Contestant * Current Folds - Starting Folds - Tickets = ATWE Total

1 * ATWE Terrahawk * 183,214,426 - 465,147 - 0 = 182,749,279
2 * ATWE Robor * 201,751,123 - 38,230,740 - 15,820,786 = 147,699,597
3 * ATWE LastKnight * 90,331,706 - 3,483,399 - 0 = 86,848,307
4 * ATWE Insidious * 141,064,961 - 58,684,902 - 681,901 = 81,698,158
5 * ATWE ReaganCRW * 73,980,336 - 10,501,161 - 0 = 63,479,175
6 * ATWE bfour * 86,888,857 - 22,005,703 - 3,077,288 = 61,805,866
7 * ATWE cbcrus * 73,031,078 - 28,117,719 - 1,414,515 = 43,498,844
8 * ATWE bob970 * 29,846,420 - 1,243,892 - 94,396 = 28,508,132
9 * ATWE FifthFactor * 28,240,319 - 684,121 - 235,000 = 27,321,198
10 * ATWE bunker * 28,231,866 - 830,790 - 958,530 = 26,442,546
11 * ATWE m0ti * 66,584,479 - 40,222,221 - 0 = 26,362,258
12 * ATWE muttley * 62,267,302 - 38,759,122 - 540,880 = 22,967,300
13 * ATWE GhettoFob * 24,352,566 - 6,169,633 - 574,898 = 17,608,035
14 * ATWE CrazySaint * 19,102,856 - 2,857,131 - 724,626 = 15,521,099
15 * ATWE Builderz * 17,262,394 - 2,429,825 - 200,000 = 14,632,569
16 * ATWE joinT * 14,255,947 - 63,313 - 1,012,653 = 13,179,981
17 * ATWE Mloot * 20,280,689 - 7,389,298 - 0 = 12,891,391
18 * ATWE TwilightZone77 * 27,034,665 - 15,566,340 - 0 = 11,468,325
19 * ATWE Rebel Alliance * 5,095,000 - 5,000 - 80,000 = 5,010,000
20 * ATWE ken008 * 7,375,034 - 1,898,137 - 487,351 = 4,989,546

This new protien is definitely slower! I remember Howard trying a different algorithm that sped our last one up... looks like we are back to the original (or some other new process)

I did snag a benchie:

Old protien (108):
------------------------------------------------------------
Distributed Folding Windows dfGUI v2.1 Benchmark

Sample Size : 119160 structures over 31224 seconds.
Protein Size: 108AA

Structures Per Second: 3.82
Structures Per Minute: 229.0
Structures Per Hour : 13739
Structures Per Day : 329728

OS : Windows XP MHz: 1775
CPU: AMD Athlon(tm) XP 2100+
------------------------------------------------------------

New protien (96)

------------------------------------------------------------
Distributed Folding Windows dfGUI v2.1 Benchmark

Sample Size : 4125 structures over 1559 seconds.
Protein Size: 96AA

Structures Per Second: 2.65
Structures Per Minute: 158.8
Structures Per Hour : 9525
Structures Per Day : 228608

OS : Windows XP MHz: 1774
CPU: AMD Athlon(tm) XP 2100+
------------------------------------------------------------

Significantly slower

Have a great night everyone!

-Sid

(and a special thanks to those who wished me well yesterday )

 

[bunker]

Woooohoooo!!!! Back ahead of m0ti!!!

I'm sure as soon as all my clients are updated I won't be though

 

[Deeko]

If my stupid computer stops randomly restarting...my production will be signifgantly higher.

 

[FoBoT]

i know its early, but still gotta give a w00t! to muttley!

#4. ATWE muttley Team AnandTech Team AnandTech 8.96

 

[CrazySaint]

Not only does is it slower, but the initial RMS's seem higher, as well.

 

[GhettoFob]

Thanks m0ti! This protein does seem slower.

 

[Robor]

A quick check of KDFold shows 1 of my 13 local clients didn't take the protein change too well. I couldn't remotely restart the service through Computer Management but a reboot of the system appears to have taken care of the problem. I've got another 11 clients at another branch. Of those 9 appear to be working and 2 are offline. Hopefully the 2 that aren't online will update okay. If so I'll be happy with the changeover. Checking Dyyryath's stats page my production is lower than the previous protein but has gone up from the overnight numbers.

ATWE Robor hits 200M folds!

WooHoo! A big milestone for me! (I know, such shamelessness)

Big congrats to the other Milestone Makers & POD's and thanks to m0ti and Sid!

 

[muttley]

could somebody give me some type of comparison number for a XP 1700+

I need to figure out how a dual proceessor is doing since so much memory is in use compared to other DC programs.

Something like timing how long it takes to get to say 80,000 folds will work fine.

muttley

 

[m0ti]

muttley,

I'm running a AXP1600+@1900+ (1400@1606 MHz).

That puts my RAM at 306 instead of 266.

I'm getting about 2.4 as opposed to 3.5 on the previous protein.

 

[muttley]

-Thanks moti

but I need like the time to get to say 10,000 or 100,000 or something.

on a XP 1700+ processor.

muttley