Folding@home Systems Synopsis

[Cliptin]

Compilation of protienA completion times
The time in minutes it takes for 1/100 of ProteinA (5/100 points) and dd:hh:mm for the whole protein.

By ID:


Cougtek

500Mhz Athlon ----------------> ~34 minutes = 56:40/5 points
800Mhz Athlon Thunderbird --> ~15.35 minutes = 25:35/5 points
XP1500+ @ 1436---------------> ~9.5 minutes = 15:50/5 points
1400 Athlon Thunderbird--------------->unknown
PII 333--------------->unknown
PIII 733--------------->unknown

P5-133XL

MP1900+ -----------------------> ~7.92 minutes = 13:12/5 points

Rocco

XP1800+ @1600 ---------------> ~9.17 minutes = 15:17/5 points

Handruin

1200 Athlon Thunderbird ---------------> ~11 minutes = 18:20/5 points
PIII 1000--------------->unknown
PIII 1000--------------->unknown
Quad PIII Xeon 500--------------->unknown

Cliptin

PI 133 (NT4)---------------> ~218.13 minutes = 15:03:23/5 points
Celeron 600 (98)---------------> ~31.73 minutes = 52:53/5 points
PIII 850 (2k)---------------> ~22.13 minutes = 36:53/5 points
Duron 850 (2k)---------------> ~15.47 minutes = 25:47/5 points
XP1700+ @1542 (2k)---------------> ~9.68 minutes = 16:08/5 points

Tea

Athlon XP 1900 (95)--------------->unknown
Athlon XP 1800 (98)--------------->unknown
Thunderbird 1400C --------------->unknown
Thunderbird 1333 --------------->unknown
Thunderbird 1000A --------------->unknown
Pentium III 666 --------------->unknown
Duron 600 --------------->unknown
K6-III 500 (98)--------------->unknown
K6-III 500 (NT4)--------------->unknown

The JoJo

Athlon XP 1800 1533@1633MHz  --------------->unknown
Athlon XP 1700  --------------->unknown
Celeron 500  --------------->unknown
Athlon 1400MHz  --------------->unknown
Celeron 600  --------------->unknown
Athlon 1200MHz  --------------->unknown
P3 733MHz  --------------->unknown
P3 666MHz  --------------->unknown
Athlon 1400Mhz  --------------->unknown
P2 333 @ 375  --------------->unknown
2-3 other computers running at my friends place --------------->unknown

Clocker

Celeron 950 (XP)--------------->unknown
700Mhz Athlon (98)--------------->unknown
PIII 1000--------------->unknown
PI 133 (95)--------------->unknown
Dual AthlonXP 1700 (XP)--------------->unknown

Prof.Wizard

1330 Athlon--------------->unknown

Bartender

1400 Athlon Thunderbird --------------->unknown
XP1600+ --------------->unknown
PIII 1200--------------->unknown
PIII 933--------------->unknown
PIII 667--------------->unknown

Sol

XP1700+ @ 1681--------------->unknown

By Processor:

MP1900+ -----------------------> ~7.92 minutes = 13:12/5 points
XP1800+ @1600 ---------------> ~9.17 minutes = 1516/5 points
XP1500+ @ 1436---------------> ~9.5 minutes = 15:50/5 points
XP1700+ @1542 (2k)---------------> ~9.68 minutes = 16:08/5 points
1200 Athlon Thunderbird ---------------> ~11 minutes = 18:20/5 points
800Mhz Athlon Thunderbird --> ~15.35 minutes = 25:08/5 points
Duron 850 (2k)---------------> ~15.47 minutes = 25:47/5 points
PIII 850 (2k)---------------> ~22.13 minutes = 36:53/5 points
Celeron 600 (98)---------------> ~31.73 minutes = 52:53/5 points
500Mhz Athlon ----------------> ~34 minutes = 56:40/5 points
PI 133 (NT4)---------------> ~218.13 minutes = 15:03:23/5 points

 

[Buck]

Compilation of protienA completion times
The time in minutes it takes for 1/100 of ProteinA (5/100 points) and dd:hh:mm for the whole protein.

By ID:

Bartender

1400 Athlon Thunderbird --------------->unknown
XP1600+ --------------->unknown
PIII 1200--------------->unknown
PIII 933--------------->unknown
PIII 667--------------->unknown

You can remove the two Athlon systems, they've been off the crunching business for about two weeks now.

 

[Cliptin]

A better formatted synopsis:

By Processor:

MP1900+ ----------------------> ~7.92 minutes = 13:12/5 points
XP1800+ @1600 ----------------> ~9.17 minutes = 1516/5 points
XP1500+ @ 1436 ---------------> ~9.5 minutes = 15:50/5 points
XP1700+ @1542 (2k) -----------> ~9.68 minutes = 16:08/5 points
1200 Athlon Thunderbird ------> ~11 minutes = 18:20/5 points
800Mhz Athlon Thunderbird ----> ~15.35 minutes = 25:08/5 points
Duron 950 (2k) ---------------> ~15.47 minutes = 25:47/5 points
PIII 850 (2k) ----------------> ~22.13 minutes = 36:53/5 points
Celeron 600 (98) -------------> ~31.73 minutes = 52:53/5 points
500Mhz Athlon ----------------> ~34 minutes = 56:40/5 points
PI 133 (NT4) -----------------> ~218.13 minutes = 15:03:23/5 points

Do we have anymore MPs around? They seem to be peppy beyond their clock speed!

PS I also corrected a typo: the Duron in the list is a 950 not 850.

 

[CNN Sport]

"I think we can help there, Dave."

"Sure can, Ed. My spies tell me that Tannin's K6-III/500s crunches one one-hundredth of a ProteinA in 50 minutes and 20 seconds, or thereabouts."

"And Tea's menagerie?"

"Sorry Ed, Tea has some pretty tight security, hard to get a camera in there. Only figure we have is for her Pentium III/666 with integrated video - a VIA chipset of some kind. But it's an interesting one: 1hour 20 minutes."

"That's a ProteinA, Dave?"

"An Ag29a Ed, but it should be close enough to the same number, I think. Sure seems slow though."

 

[Bartender]

Dave, all three of my Pentium III systems seem slow at this crunching game.

 

[Cliptin]

Crunching is faster on a more simplistic OS as well. Did CNN sport happen to spot which OS the "50 minutes and 20 seconds" could be attibuted to?

 

[CougTek]

With the new core v.2.47, it takes 33m42s to crunch 1/100th of a proteinA on my Athlon 500MHz under Linux (Red Hat 7.3).

You can remove the Pentium III 733MHz from my list of computers.

Unfortunately, I haven't had the chance to get a proteinA on my Athlon XP @ 1436MHz recently so I can't give you a more precise number. My 9½ minutes estimate seems a bit optimistic. I hope I'll have the chance to confirm/change it soon.

 

[CNN Sport]

"Dave?"

"Yes Ed. The 50:20 score on the K6-III was using Win98SE. I gather that Tannin has another K6-III as well, running Windows NT, but we could only get timings for some other protein on that one. Guess we will have to wait for game day to learn more."

"And Tea's Pentium III?"

"98SE also, Ed."

 

[P5-133XL]

You can add a 266P2 takes 68.8 minutes for 1/100 of a proteinA for a 114.7 hours/5Pts

 

[Clocker]

Do i just have to watch to see if my system happens to be running a ProteinA in order to report the result?

C

 

[Clocker]

Protein: GAAA_tetraloop (I'm doing tons of these things lately)

Athlon 1700+ WinXP: 3:45 (225 seconds per %)
P3 1Ghz WinXP : 7:30 (450 seconds per %)

C

 

[CougTek]

As a comparison, my Atlon XP does 1/100th of a GAAA_blablabla (same as Clocker wrote) in 3:58. 13 seconds more, but that system is also my main computer so there are often other applications (mainly browsers) running at the same time.

 

[Clocker]

As a comparison, my Atlon XP does 1/100th of a GAAA_blablabla (same as Clocker wrote) in 3:58. 13 seconds more, but that system is also my main computer so there are often other applications (mainly browsers) running at the same time.

Same here Coug...my Athlon system is running all kinds of stuff....Mail, Remote Desktop (to control the P3 system) , Browsers, Counter-Strike, NAV, IM, Counter Strike, Counter Strike, Counter Strike...

 

[James]

Any Counterstrike, there, Clocker? :mrgrn:

 

[The JoJo]

Athlon XP 1800 1533@1633MHz --------------->unknown
Celeron 500 --------------->unknown
Celeron 600 --------------->unknown
Athlon 1200MHz --------------->unknown
P3 733MHz --------------->unknown
P3 666MHz --------------->unknown
Athlon 1400Mhz --------------->unknown
P2 333 @ 375 --------------->unknown

I'm down to these systems at the moment . I think only one or two are not 24/7.

I'll try to check some times for the protein a.

We are rocking! Top 100, yess!

 

[Adcadet]

well, I can add my dual 1.2GHz Athlon MPs if that will help. Just give me a day or so to set it up.

 

[Jake the Dog]

when my faulty athlon 2100+ is replaced, i'll join up as well. this should be early next week. i plan to overclock it to 2200+ so hopefully it'll rock at folding.

 

[Prof.Wizard]

Are you sure it it takes the same amount of time for each 1/100 of ProteinA?

I haven't yet figured how time it takes to do 1% and if this is done in stable periods of time... :-?

 

[CougTek]

Great. Thanks for joining guys.

 

[The JoJo]

I'm glad you can join us in the crunching guys! That will certainly help us climb a few steps higher in the ranking

 

[Clocker]

Yeah! It's about time you got in on the action, Andy!!

C

 

[Sol]

My PC has undergone some serious changes in the last couple of days so the new specs are:

AthlonXP 1900+ @ 1669Mhz-------> ~8min/frame for Protein A

This will be invalidated either tomorrow or monday though when I overclock my PC a bit more.

 

[Cliptin]

Are you sure it it takes the same amount of time for each 1/100 of ProteinA?

I haven't yet figured how time it takes to do 1% and if this is done in stable periods of time... :-?

A: NO, I am not sure.

B:It will be close enough. This is not part of the competition.

 

[Cliptin]

Oops, I did not mean to be shouting at you Prof.

My PIII will do a frame of GAAA_tetraloop in 8:58.
My PIII will do a frame of ProteinA in 22.13.

This is a statistically significant difference.

I also suspect that it takes slightly longer for an Ag29a though I need to evaluate it more closely.

FWIW, The program FoldMonitor will keep a running average for whatever protein you are currently working on. It works on linux too.

These stats might be fun to look at.

An alternative method for setting up F@hv3 as a service. And for those of you who just want to hide to window there is this.

PS. Thanks for your updates everyone.

 

[Handruin]

Ok, ok, we have so much activity that I will try and get a folding page going for members soon for us to enter all sorts of info.

Any ideas what would be a good starting ground for features? Should I start a new topic?

 

[Adcadet]

My PIII will do a frame of GAAA_tetraloop in 8:58.
My PIII will do a frame of ProteinA in 22.13.

This is a statistically significant difference.

Based upon what test?

 

[Cliptin]

My PIII will do a frame of GAAA_tetraloop in 8:58.
My PIII will do a frame of ProteinA in 22.13.

This is a statistically significant difference.

Based upon what test?

Some folks are looking at the console window and subtracting times. I am using FoldMonitor which averages the times for the frames completed so far.

 

[Cliptin]

Ok, ok, we have so much activity that I will try and get a folding page going for members soon for us to enter all sorts of info.

Any ideas what would be a good starting ground for features? Should I start a new topic?

A new thread would help keep this one as clean as possible. I would help you build a page.

 

[CougTek]

Update :

I finally got a ProteinA on my Athlon XP 1.436GHz and with the latest core v2.47 and when I let it work quietly (ie, not MSNing and surfing), it completes 1/100th in 8 minutes, 43 secondes (average of the first 7 segments). Not bad.

 

[Prof.Wizard]

For me every percentage of the proteinA is done in different amount of times. From 9 to 18 mins.

Just to remind you, I have an Athlon T-bird clocked @ 1343MHz.

 

[CougTek]

It's probably because you use your system a lot while the client is running. For instance, when I start Serious Sam, the client completly stops processing proteins until I quit the game (or at least that's what the Genome@home client did). Don't forget that F@h only use the cycles that no other applications use (unless you change its priority).

 

[Clocker]

That fold monitor program is cool:

 

[Prof.Wizard]

Where do I get it C?

 

[Clocker]

http://users.neosmart.com/hruzaden/

 

[NRG = mc²]

Indead, it's probably a Katmai or a similar Xeon version. That's about the same reflection I had.

If its a Xeon, it will have full speed cache. Just thought I'd point that out.

 

[NRG = mc²]

Whoops scrap that. Confused the windows...

What I meant to say was:

My XP is now off for the summer (about a week ago I put it in the attic). From now till September, I've got a useless P3-733. I'm going to get my butt kicked.

 

[Prof.Wizard]

My XP is now off for the summer (about a week ago I put it in the attic). From now till September, I've got a useless P3-733. I'm going to get my butt kicked.

That's right Will... :wink:

But what the... I nearly forgot: my computer will be off for all August too!!
I guess we have to stop these trips to Greece William, don't we?!
They "hurt" our standings...

 

[P5-133XL]

because of the UD upgrade, the amount of CPU wastage has gone drasticly up the last week. To compensate, I've added F@H processes to some machines so that at least something is running while UD is waiting to connect.

I don't think it will add alot of WU's because these are slow machines, but it is better than nothing.

 

[Pradeep]

I guess we have to stop these trips to Greece William, don't we?!
They "hurt" our standings...

Why not let it keep running whilst you are away? If I recall you have broadband right?

 

[Prof.Wizard]

Why not let it keep running whilst you are away? If I recall you have broadband right?

I'm not living a 80'C grads Athlon box working in Rome's August Pradeep... Not even for such a valiant cause...