PDA

View Full Version : CASP5 Protein #2 Benchmarks



MAD-ness
06-11-2002, 06:10 PM
Now that the protein change-over has occured, I figured it is time to start a thread where benchmarks for this protein can be submitted.

If someone feels like doing something fancy with benchmarks, it would be greatly appreciated but until they do, I will collect them by hand and toss them up on a page. Here is a sample from the last protein:

http://www.geocities.com/m4dn3ss2000/DFBenchResultsTable.html

You might get pop-ups or other annoying crap there, I am not sure.

I will try to get the benchmarks for this protein up before the end of the week and I should be able to put them at www.teamstirfry.com so that you don't have to deal with Yahoo/GeoShitties.

Please run the new protein for several hours (atleast two or so) and preferably run it overnight and submit them tommorow.

Please include information from your system that the benchmark script might not include, such as processor model (especially with P3s and Celerons in different models), cache size and fsb (if these vary for a particular CPU, like the aforementioned P3 lines).

Thanks.

Digital Parasite
06-12-2002, 07:50 AM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.1 Benchmark

Sample Size: 10100 structures over 35640 seconds.

Structures Per Second: 0.28
Structures Per Minute: 17.1
Structures Per Hour : 1024
Structures Per Day : 24565

OS : Microsoft Windows XP MHz: 801
CPU: GenuineIntel x86 Family 6 Model 8 Stepping 3
------------------------------------------------------------
^^^ P3-800

------------------------------------------------------------
Distributed Folding Linux Benchmark Script V1.0

Sample Size: 20955 structures over 61815 seconds.

Structures Per Second: 0.34
Structures Per Minute: 20.34
Structures Per Hour: 1220.38
Structures Per Day: 29289.12

Linux OS - Running Kernel Version 2.4.17
Celeron (Coppermine) @ 908mhz (128 KB cache)
------------------------------------------------------------

PackSwede
06-12-2002, 10:18 AM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.1 Benchmark

Sample Size: 42905 structures over 76473 seconds.

Structures Per Second: 0.56
Structures Per Minute: 33.7
Structures Per Hour : 2020
Structures Per Day : 48475

OS : Microsoft Windows XP MHz: 1395
CPU: Intel(R) Pentium(R) 4 CPU 1400MHz
------------------------------------------------------------

Almost double the production compared to the 182AA Beast :cool:

/PackSwede

Angus
06-12-2002, 10:20 AM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.1 Benchmark

Sample Size: 12725 structures over 48562 seconds.

Structures Per Second: 0.26
Structures Per Minute: 15.7
Structures Per Hour : 943
Structures Per Day : 22643

OS : Windows 98 MHz: 707
CPU: GenuineIntel x86 Family 6 Model 8 Stepping 3
------------------------------------------------------------
Celeron 566 OC to 708, 128MB PC100

Digital Parasite
06-12-2002, 10:20 AM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.1 Benchmark

Sample Size: 45700 structures over 74254 seconds.

Structures Per Second: 0.62
Structures Per Minute: 36.9
Structures Per Hour : 2216
Structures Per Day : 53178

OS : Microsoft Windows 2000 MHz: 1495
CPU: Intel(R) Pentium(R) 4 CPU 1.50GHz
------------------------------------------------------------

jdhaenens
06-12-2002, 11:32 AM
Laptop...400mhz celeron, 128MB SDRAM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.0 Benchmark

Sample Size: 1410 structures over 9301 seconds.

Structures Per Second: 0.15
Structures Per Minute: 9.1
Structures Per Hour : 546
Structures Per Day : 13099

OS : Windows 98 MHz: 397
CPU: GenuineIntel Intel(r) Celeron(tm) Processor



Wife's Computer ...800MHZ Duron, 128MB SDRAM

------------------------------------------------------------
Distributed Folding Windows dfGUI v1.0 Benchmark

Sample Size: 3050 structures over 9484 seconds.

Structures Per Second: 0.32
Structures Per Minute: 19.3
Structures Per Hour : 1159
Structures Per Day : 27815

OS : Windows 98 MHz: ???
CPU: AuthenticAMD AMD Duron(tm) processor
------------------------------------------------------------

My Computer... Dual Athlon MP-2000 1GB ECC reg DDR

CPU0

------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 5550 structures over 10174 seconds.

Structures Per Second: 0.55
Structures Per Minute: 32.8
Structures Per Hour : 1970
Structures Per Day : 47276

OS : Microsoft Windows XP MHz: 1680
CPU: AMD Athlon(TM) XP 1900+
------------------------------------------------------------

CPU1

------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 5550 structures over 10167 seconds.

Structures Per Second: 0.55
Structures Per Minute: 32.9
Structures Per Hour : 1976
Structures Per Day : 47416

OS : Microsoft Windows XP MHz: 1680
CPU: AMD Athlon(TM) XP 1900+
------------------------------------------------------------

Angus
06-12-2002, 11:47 AM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.1 Benchmark

Sample Size: 57475 structures over 76018 seconds.

Structures Per Second: 0.76
Structures Per Minute: 45.4
Structures Per Hour : 2722
Structures Per Day : 65325

OS : Microsoft Windows 2000 MHz: 1993
CPU: Intel(R) Pentium(R) 4 CPU 2.00GHz
------------------------------------------------------------
P4 2GHz 512MB RDRAM

Nitrousine
06-12-2002, 01:46 PM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.0 Benchmark

Sample Size: 23600 structures over 49905 seconds.

Structures Per Second: 0.47
Structures Per Minute: 28.4
Structures Per Hour : 1702
Structures Per Day : 40858

OS : Microsoft Windows 2000 MHz: 1025
CPU: AMD Duron(tm) processor
------------------------------------------------------------

128MB PC133 SDRAM

Paratima
06-12-2002, 07:02 PM
Donald: 256MB - DDR2100
Distributed Folding Windows dfGUI v1.1 Benchmark
Sample Size: 60350 structures over 91554 seconds.

Structures Per Second: 0.66
Structures Per Minute: 39.6
Structures Per Hour : 2375
Structures Per Day : 56992

OS : Microsoft Windows 2000 MHz: 1587
CPU: AMD Athlon(tm) XP 1900+ (actually an 1800 lightly OC'd)
------------------------------------------------------------
Daisy: 256MB - DDR2100
Distributed Folding Windows dfGUI v1.1 Benchmark
Sample Size: 46750 structures over 74587 seconds.

Structures Per Second: 0.63
Structures Per Minute: 37.6
Structures Per Hour : 2258
Structures Per Day : 54196

OS : Microsoft Windows 2000 MHz: 1465
CPU: AMD Athlon(tm) XP 1700+

MAD-ness
06-13-2002, 01:34 AM
Here are the submitted entries in a chart:

http://www.geocities.com/m4dn3ss2000/CASP5Protein2.html

I guessed on the Duron cache size since I can't remember and I might have goofed a few other details on architectures, as always, when I had to guess (or was just brain-dead).

Could definitely use some more submissions.

Gonna see if I can talk jdhaenens into running his again, he has interesting hardware but his sample set was relatively small (maybe an hour or so of run time) and if I am going to manually enter this data for the moment, I wanna make sure it is pretty accurate.

Anyone have anything interesting to submit that isn't really "covered" by benches already submitted?

A highly overclocked P4, a highly overclocked AthlonXP, a couple Tbirds, a classic Athlon or two, and a tualatin P3 would all be nice. Haven't seen the ubiquitous Celeron300a @ 450 yet either on this protein.

collin
06-13-2002, 03:18 AM
Pentium III 450 (4.5x100), 512k L2 [Katmai]
WinNT 4.0SP6 (~449Mhz)
256Mb PC100
Running as a service
-----
Structures Per Second: 0.21
Structures Per Minute: 12.6
Structures Per Hour : 758
Structures Per Day : 18202

Sample size : 21850 structures

======


Pentium III 866 (6.5x133),256k L2 [Coppermine]
WinNT 4.0SP6 (~864Mhz)
192Mb PC133
Running as a service
-----
Structures Per Second: 0.34
Structures Per Minute: 20.3
Structures Per Hour : 1220
Structures Per Day : 29281

Sample size : 35100 structures

=====


Pentium III 866 (6.5x133),256k L2 [Coppermine]
Win2K (~863Mhz)
128Mb PC133
Running as a service
-----
Structures Per Second: 0.32
Structures Per Minute: 19.3
Structures Per Hour : 1160
Structures Per Day : 27851

Sample size : 32350 structures

Welnic
06-13-2002, 03:20 AM
first df client
Sample size: 10178 structures over 36720 seconds

Structures per minute: 16.6
Structures per hour: 998
Structures per day: 23948

second df client
Sample size: 10012 structures over 36900 seconds

Structures per minute: 16.3
Structures per hour: 977
Structures per day: 23443

OS: OS 10.1.5
CPU dual 800MHz G4

Not the most cost effective box for farming.

Eaglechild
06-13-2002, 04:23 AM
iBook 600mhz (notebook)
Mac OS-X 256k L2 cache
per sec: .158
per min: 9.5
per hour: 570
per day: 13,680

Power Mac 800mhz (single proc)
Mac OS-X 256k L2 cache
per sec: .23
per min: 14
per hr: 845
per day: 20280

Celeron 1.3 ghz
Windows XP 256k L2 cache
per sec: .319
per min: 19.16
per hr: 1150
per day: 27600

Athlon 1900+
Mandrake Linux
per sec: .56
per min: 33.66
per hr: 2020
per day: 48,480

jdhaenens
06-13-2002, 08:18 AM
Mad-ness Here they are:

Dual Athlon MP-2000 1GB ECC reg DDR (PC 2100)

CPU 0

------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 47400 structures over 84820 seconds.

Structures Per Second: 0.56
Structures Per Minute: 33.6
Structures Per Hour : 2013
Structures Per Day : 48315

OS : Microsoft Windows XP MHz: 1680
CPU: AMD Athlon(TM) XP 1900+

CPU 1

Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 46950 structures over 84807 seconds.

Structures Per Second: 0.55
Structures Per Minute: 33.2
Structures Per Hour : 1994
Structures Per Day : 47845

OS : Microsoft Windows XP MHz: 1680
CPU: AMD Athlon(TM) XP 1900+

I'd overclock it but for every minute I spend tweaking, I'd lose at least 34 folds (not counting the shutdown and restart time).

Jim

Paratima
06-13-2002, 09:03 AM
Huey, Dewey, Louie, & Grandma: Dell wkstations
Distributed Folding Windows dfGUI v0.6 Benchmark
Sample Size: 2600 structures over 5431 seconds.
Structures Per Second: 0.49
Structures Per Minute: 29.2
Structures Per Hour : 1752
Structures Per Day : 42052
OS : Microsoft Windows 2000 MHz: 1495
CPU: Intel(R) Pentium(R) 4 CPU 1.50GHz
------------------------------------------------------------
Scrooge: Athlon Classic - SuSE Linux
Thu Jun 13 07:21:24 EDT 2002
Distributed Folding Linux Benchmark Script V1.1
Sample Size: 49640 structures over 145761 seconds.
Structures Per Second: 0.34
Structures Per Minute: 20.43
Structures Per Hour: 1226.01
Structures Per Day: 29424.24
Linux OS - Running Kernel Version 2.4.10-4GB
AMD Athlon(tm) Processor @ 848mhz (512 KB cache)

collin
06-13-2002, 10:49 AM
More results :

Athlon XP 1500+ (10x163)
WinME
256Mb DDR266
Console version off commandline
-----
Structures Per Second: 0.75
Structures Per Minute: 45.0
Structures Per Hour : 2700
Structures Per Day : 64780

Sample size : 120200 structures

======


Athlon Thunderbird 1.0G (10x153)
Win98SE
256Mb PC133
Console version off commandline
-----
Structures Per Second: 0.67
Structures Per Minute: 40.0
Structures Per Hour : 2402
Structures Per Day : 57640

Sample size : 109300 structures

MAD-ness
06-13-2002, 06:53 PM
Thanks for the extra submissions guys. Some interesting boxen there. :)

The G4s seem to do about the same per/cycle as the P3 just from glancing at the submitted benchmarks. The Macs don't scale to the same sort of frequencies as the K7 and the P4 cored chips, but considering how poorly the Macs do on some DC projects, those results don't look too bad to me. Not like RC5 results on them, of course, but not horrible either.

If my DSL stays up tonight I will add these submissions to the page as well as trying to get the page put up on a 'real' web site. I just added in my 3 machines and with the ones just submitted that should give us ~20 for this protein. :)

GOLDENBALLSAINTYORK
06-14-2002, 12:05 PM
384 Mb RAM 100MHz

MashRinx
06-14-2002, 05:29 PM
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 126808 structures over 165279 seconds.

Structures Per Second: 0.77
Structures Per Minute: 46.0
Structures Per Hour : 2762
Structures Per Day : 66289

OS : Microsoft Windows XP MHz: 2193
CPU: Intel(R) Pentium(R) 4 CPU 2.20GHz
------------------------------------------------------------

MashRinx

MAD-ness
06-14-2002, 08:16 PM
ok, added in new benchmarks.

Forgot to add in jdhaeden's laptop and I think someone at Ars posted a K6-3 bench and I have had my celeron266 running for 24 hours, but oh well, next protein I will grab all them, the info is here for anyone who is really interested anyways.

Fastest bench submitted here was MashRinx's P4 2.2ghz Northwood.

Collin's two heavily overclocked AMD CPUs had great results though. I don't know if I got the stock fsb/multiplier info right on those cpus though Collin. I guessed on the XP1500+ frequency since it is one of the XP models I do not know in terms of frequency. I assumed that you left the multipliers locked and went for straight FSB overclocking.

The processors that have "(1)" or "(2)" in the Processor name were the first and second cpus of duallies. This isn't the most efficient way of doing this, but I thought the info on how the second cpu performs (reallly close to the first CPU and not much slower than a single CPU set-up) would be useful.

http://www.geocities.com/m4dn3ss2000/CASP5Protein2.html

I have sent off the file to pointwood and we will be getting it up on www.teamstirfry.com but I don't know jack about HTML or any other web nonsense (or computers in general) so I don't know exactly how long it will take for us to play email tag over about 10 timezones and get it set up.

Thanks for the submissions, I will try to get it up faster next week, I had a ton of real life issues this week not to mention no internet access for a while. :(

MAD-ness
06-14-2002, 08:29 PM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.1 Benchmark

Sample Size: 9149 structures over 88447 seconds.

Structures Per Second: 0.10
Structures Per Minute: 6.2
Structures Per Hour : 372
Structures Per Day : 8938

OS : Windows 98 MHz: ???
CPU: GenuineIntel Intel(r) Celeron(tm) Processor
------------------------------------------------------------

Paratima
06-14-2002, 08:35 PM
Hey, thanks, MAD-ness! You're doing a good service here. :thumbs:

I've already made a couple of changes based on the numbers I've seen and just from the exercise of running the numbers on my own boxen. Very helpful.

Also toots on the horn for Dyyryath for the Perl script and Digital Parasite for dfGUI, so we can get these measurements to begin with!

fimmi
06-15-2002, 06:11 AM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.1 Benchmark

Sample Size: 154475 structures over 155740 seconds.

Structures Per Second: 0.99
Structures Per Minute: 59.5
Structures Per Hour : 3571
Structures Per Day : 85698

OS : Microsoft Windows XP MHz: 1579
CPU: AMD Athlon(tm) XP 1900+
------------------------------------------------------------


XP 1600+ FSB overclock
FSB 150 MHz

jdhaenens
06-15-2002, 07:35 AM
New Benchmarks after slight FSB overclocking (138 MHZ)...the second service refuses to run above 140 MHZ FSB.

A7M266-D Mobo dual Athlon MP-2000 1GB ECC Reg DDR Ram

CPU0

------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 13650 structures over 23331 seconds.

Structures Per Second: 0.59
Structures Per Minute: 35.2
Structures Per Hour : 2111
Structures Per Day : 50662

OS : Microsoft Windows XP MHz: 1725
CPU: AMD Athlon(TM) XP 1900+
------------------------------------------------------------


CPU 1

------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 13800 structures over 23328 seconds.

Structures Per Second: 0.59
Structures Per Minute: 35.5
Structures Per Hour : 2130
Structures Per Day : 51111

OS : Microsoft Windows XP MHz: 1725
CPU: AMD Athlon(TM) XP 1900+
------------------------------------------------------------

Digital Parasite
06-15-2002, 08:03 AM
------------------------------------------------------------
Distributed Folding Windows dfGUI v1.1 Benchmark

Sample Size: 23950 structures over 32957 seconds.

Structures Per Second: 0.73
Structures Per Minute: 43.6
Structures Per Hour : 2617
Structures Per Day : 62820

OS : Microsoft Windows XP MHz: 2259
CPU: Intel(R) Pentium(R) 4 CPU 2.26GHz
------------------------------------------------------------

MashRinx
06-15-2002, 12:52 PM
Hey MAD-ness, would you like some lower level Benchmarks too? I have a few slower machines set up in a closet at home that are crunching away. P2-233's, my wife's using a P2-400, etc.

Also, at work I'm running on a couple P3-XEONs. I dont know if there's been talk in the past on wether the extra cache makes a different (I haven't been around D.C. much in a while), bu I could submit those if it would help too....

MashRinx (Kevin)

MAD-ness
06-15-2002, 01:30 PM
Actually, MashRinx, I would LOVE all of those benchmarks if at all possible.

Someone has asked me about P1/P1MMX CPUs and I only have a P1 200mhz here that I have access to (and keep forgetting to benchmark).

If I ever figure out what chipset this Celeron266 box I was given has, I will hopefully be able to update it to some sort of P2, so low end P2 numbers would help me personally.

A lot of people have access to older machines but are not sure if it is worth the effort to set them up. Since DF is pretty much a frequency dependent program, my opinion so far is that if the machines are already running, it is definitely worthwhile. Unlike SETI, for example, you aren't going to murder your average WU time and the production scales pretty well backwards through architectures (i.e. the performance scales per clock cycle in both directions, reasonably well). Data to confirm this would be great.

The Xeons...yum. :) Would really like to see those large cache chips in action. =)

MashRinx
06-16-2002, 08:11 PM
Unfortunately, I can only use these machines for DF until later tonight. This is a mojor DB application used by our company, and there's NO way I am going to run D.C. on these boxes after we get everything up and running, and that will happen early tomorrow AM..... At least I was able to "burn them in" over the weekend, and get some good benchmarks too..... :)


Database Server:
Dell PowerEdge 6400 - Quad Capable machine w/ 2 Procs
Both Xeon 700's with 2MB of Cache
1 GB RAM


Processor #1:
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 45 structures over 129 seconds.

Structures Per Second: 0.35
Structures Per Minute: 20.9
Structures Per Hour : 1256
Structures Per Day : 30140

OS : Microsoft Windows 2000 MHz: 699
CPU: GenuineIntel x86 Family 6 Model 10 Stepping 4
------------------------------------------------------------


Processor #2:
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 35 structures over 130 seconds.

Structures Per Second: 0.27
Structures Per Minute: 16.2
Structures Per Hour : 969
Structures Per Day : 23262

OS : Microsoft Windows 2000 MHz: 699
CPU: GenuineIntel x86 Family 6 Model 10 Stepping 4
------------------------------------------------------------



Application Servers:


App Server #1:
Compaq Proliant ML570 - Quad CPU Machine w/ 2 CPU Installed
Processors: Xeons w/ 2MB Cache
1 GB Ram


Processor #1:
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 35 structures over 194 seconds.

Structures Per Second: 0.18
Structures Per Minute: 11.1
Structures Per Hour : 663
Structures Per Day : 15916

OS : Microsoft Windows 2000 MHz: 701
CPU: GenuineIntel x86 Family 6 Model 10 Stepping 4
------------------------------------------------------------


Processor #2:
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 55 structures over 206 seconds.

Structures Per Second: 0.27
Structures Per Minute: 16.0
Structures Per Hour : 961
Structures Per Day : 23068

OS : Microsoft Windows 2000 MHz: 701
CPU: GenuineIntel x86 Family 6 Model 10 Stepping 4
------------------------------------------------------------



App Server #2:
Compaq Proliant ML570 - Quad CPU Machine w/ 2 CPU Installed
Xeons again....
1 GB Ram


Processor #1:
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 30 structures over 101 seconds.

Structures Per Second: 0.38
Structures Per Minute: 22.5
Structures Per Hour : 1350
Structures Per Day : 32400

OS : Microsoft Windows 2000 MHz: 701
CPU: GenuineIntel x86 Family 6 Model 10 Stepping 4
------------------------------------------------------------


Processor #2:
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 25 structures over 94 seconds.

Structures Per Second: 0.27
Structures Per Minute: 16.1
Structures Per Hour : 968
Structures Per Day : 23226

OS : Microsoft Windows 2000 MHz: 701
CPU: GenuineIntel x86 Family 6 Model 10 Stepping 4
------------------------------------------------------------


Reporting Servers:


Reporting Server #1:
P4 - 1.2
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 15 structures over 33 seconds.

Structures Per Second: 0.48
Structures Per Minute: 29.0
Structures Per Hour : 1742
Structures Per Day : 41806

OS : Microsoft Windows 2000 MHz: 1125
CPU: Intel(R) Pentium(R) III CPU family 1133MHz
------------------------------------------------------------


Reporting Server #2:
P4 - 1.2
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 20 structures over 56 seconds.

Structures Per Second: 0.43
Structures Per Minute: 25.5
Structures Per Hour : 1532
Structures Per Day : 36766

OS : Microsoft Windows 2000 MHz: 1125
CPU: Intel(R) Pentium(R) III CPU family 1133MHz
------------------------------------------------------------


Hope these help....

MashRinx

MashRinx
06-16-2002, 08:15 PM
The current "home folding fleet" :


P2-400 w/ approx 128 MB RAM
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 9 structures over 46 seconds.

Structures Per Second: 0.20
Structures Per Minute: 12.3
Structures Per Hour : 736
Structures Per Day : 17673

OS : Microsoft Windows .NET MHz: 398
CPU: GenuineIntel x86 Family 6 Model 5 Stepping 2
------------------------------------------------------------


DFGUI doesn't pick it up, but this is a P2-233
64 MB RAM
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 95 structures over 797 seconds.

Structures Per Second: 0.12
Structures Per Minute: 7.3
Structures Per Hour : 438
Structures Per Day : 10510

OS : Windows Millennium Edition MHz: ???
CPU: GenuineIntel Pentium(r) II Processor
------------------------------------------------------------


DFGUI doesn't pick it up, but this is a P2-233
64 MB RAM
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 29 structures over 586 seconds.

Structures Per Second: 0.05
Structures Per Minute: 3.1
Structures Per Hour : 183
Structures Per Day : 4404

OS : Windows Millennium Edition MHz: ???
CPU: GenuineIntel Pentium(r) II Processor
------------------------------------------------------------


This box is a Celleron (sp?)
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 140 structures over 647 seconds.

Structures Per Second: 0.22
Structures Per Minute: 13.0
Structures Per Hour : 783
Structures Per Day : 18783

OS : Microsoft Windows XP MHz: 432
CPU: GenuineIntel x86 Family 6 Model 6 Stepping 5
------------------------------------------------------------



DFGUI doesn't pick it up, but this is a P2-233
64 MB RAM
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 34 structures over 315 seconds.

Structures Per Second: 0.11
Structures Per Minute: 6.6
Structures Per Hour : 396
Structures Per Day : 9507

OS : Windows 98 MHz: ???
CPU: GenuineIntel Pentium(r) II Processor
------------------------------------------------------------


P2-400 w/ 192 MB RAM
------------------------------------------------------------
Distributed Folding Windows dfGUI v0.6 Benchmark

Sample Size: 12 structures over 47 seconds.

Structures Per Second: 0.26
Structures Per Minute: 15.3
Structures Per Hour : 919
Structures Per Day : 22060

OS : Microsoft Windows XP MHz: 400
CPU: GenuineIntel x86 Family 6 Model 5 Stepping 2
------------------------------------------------------------


MashRinx

RipItUp
06-20-2002, 02:00 AM
Hello MAD-ness,

Nice idea by the way.

I thought I'd put it on my home PC's but came up against several problems. Two of them gave values over 500 000 per day ! I believe you have seen this before and know how to correct them ..please advise so I can test them again.

Another gave a result which looks suspiciously large :-


Structures Per Second: 0.86
Structures Per Minute: 51.4
Structures Per Hour : 3085
Structures Per Day : 74052 !!

OS : Microsoft Windows 2000
CPU: Athlon Tbird 1333 @ 1400


This looks a bit large !
-------------------------------------------

Structures Per Second: 0.99
Structures Per Minute: 59.5
Structures Per Hour : 3573
Structures Per Day : 85758

OS : Microsoft Windows XP
CPU: AMD XP 1900 @ 1740 ( Palamino )

This machine keeps losing DF ! It runs and then when I next look at it it is not there. No error it just stops ???!!!! :bang:

----------------------------------------------------------

Structures Per Second: 0.61
Structures Per Minute: 36.9
Structures Per Hour : 2214
Structures Per Day : 53130

OS : Microsoft Windows 2000
CPU: AMD Duron Morgan core 1200@1200

---------------------------------------------------------------


One of the machines that gives too large a result is actually the same as the 1333@1400 mentioned above, same chip and same operating system, however it differs in the RAM it has, one has DDR and one has SDRAM. IF I can get that working fine it will be nice to see the difference DDR RAM has on ouput.

Regards

Andy

RipItUp
06-20-2002, 01:04 PM
I got home today and noticed my main machine had not crapped out on me for once and had lasted all day and managed to get over 1 protein run per second !

http://www.aocb77.dsl.pipex.com/dfrate.jpg

I'm running this at a FSB of 145. Interestingly there is a result on the first page where someone reaches this sort of speed with 200MHz or so less, but they are running at a FSB of 150 !

From all the results generated does it show that FSB has a large effect on the speed of production ?

I still need to get my DDR and SDRAM boards not to give erronous values so I can see the difference this makes.

Regards


Andy

MAD-ness
06-21-2002, 06:35 PM
The DF client primarily requires clock cycles. It spends most of its time traversing a "tree-like" structure...doing pointer traversals. Both of the previous statements are paraphrasings of things Howard has stated. The Alpha is the king of performance in DF, per clock cycle, but this doesn't really say much, as the Alpha seems to be the per clock cycle performance leader in just about any app which is (well) written for it. ;)

One thing that I noticed with the last set of benchmarks is that a large L2 does not seem to make much difference.

I have noticed a big performance boost for people with higher FSB speeds (if you have a choice between raising the multiplier or the FSB speed it seems like a no-brainer for almost any app to up the FSB) but the highest submitted FSB speeds on AthlonXP CPUs with DDR SDRAM don't appear to be the highest submitted scores on CPUs in that frequency range.

I would speculate that memory timings and latencies make a significant difference, but I don't really have any solid evidence to back this up. The P4 seems to do quite well regardless of the memory architecture or speed in use.

I haven't and won't be able to collect any benches for the current protein because I have no internet connection currently, but if people post them I will, of course, add them to a page at a later date and incorporate them into whatever database I get set up eventually.

IronBits
06-22-2002, 11:08 PM
Distributed Folding Windows dfGUI v1.3 Benchmark

Sample Size: 1875 structures over 1851 seconds.

Structures Per Second: 1.01
Structures Per Minute: 60.9
Structures Per Hour : 3653
Structures Per Day : 87662

OS : Microsoft Windows 2000 MHz: 1603
CPU: AMD Athlon(tm) XP 1900+