REPORT WU-SPEEDS HERE !

log in

Advanced search

Message boards : Cruncher's Corner : REPORT WU-SPEEDS HERE !

1 · 2 · 3 · 4 . . . 10 · Next
Author Message
NIMRUTH
Avatar
Send message
Joined: 24 Feb 05
Posts: 12
Credit: 1,886
RAC: 0
Message 5876 - Posted: 27 Feb 2005, 12:06:13 UTC

Well, I`m running an OC`d sempron 2200+ @ 1710Mhz with 256Mb 2700DDR kingston on a Asrock k7s41gx...

Setispeed : aprox. 4.25 Hrs/WU

Einsteinspeed : aprox. 7.5 Hrs/WU

Lookslike Sempron kicks P4 ass..

Since I`m lookin` for an upgrade, what setup will be a major improvement ? In speed that is.. I was thinking that I`d be better off adding a "cheapo' Sempron 3100+ system, instead of going for 64bit AMD`s or the HT P4`s.. I mean how fast is a the top of the line anyway? I bet none go under the 4 Hr/wu for Einstein..

In that light, I`d get the same speed if I got a cheapo asrock and a new sempron .. that would be a lot cheaper, right?
____________

NIMRUTH
Avatar
Send message
Joined: 24 Feb 05
Posts: 12
Credit: 1,886
RAC: 0
Message 5877 - Posted: 27 Feb 2005, 12:18:25 UTC

--- - 2005-02-27 13:15:47 - Suspending computation and network activity - running CPU benchmarks
--- - 2005-02-27 13:16:48 - Benchmark results:
--- - 2005-02-27 13:16:48 - Number of CPUs: 1
--- - 2005-02-27 13:16:48 - 1592 double precision MIPS (Whetstone) per CPU
--- - 2005-02-27 13:16:48 - 3855 integer MIPS (Dhrystone) per CPU
--- - 2005-02-27 13:16:48 - Finished CPU benchmarks


is that any good?
____________

Profile ErichZann
Avatar
Send message
Joined: 11 Feb 05
Posts: 121
Credit: 81,582
RAC: 0
Message 5880 - Posted: 27 Feb 2005, 12:31:17 UTC

i got an athlon XP 2200+ at 1900 Mhz and 768 MB DDR at 266 Mhz on an Asrock K7VT2 and i takes about 3.5 hours for a Seti@Home and about 7 hours for an Einstein@Home WU.
____________

NIMRUTH
Avatar
Send message
Joined: 24 Feb 05
Posts: 12
Credit: 1,886
RAC: 0
Message 5881 - Posted: 27 Feb 2005, 12:54:37 UTC

Just tried to crank it up to 210FSB, but it keeps freezing up just as windows is starting.. damn..

I got it stable at 1800mhz, cpu running a bit hot.. 53-55 C ..Cpuidle keeps it in check, but this costs cycles..

benchmark at 1800hz

--- - 2005-02-27 13:50:47 - Benchmark results:
--- - 2005-02-27 13:50:47 - Number of CPUs: 1
--- - 2005-02-27 13:50:47 - 1679 double precision MIPS (Whetstone) per CPU
--- - 2005-02-27 13:50:47 - 4044 integer MIPS (Dhrystone) per CPU
--- - 2005-02-27 13:50:47 - Finished CPU benchmarks

might even set it a bit lower, I`m running it at the limit it seems..
____________

NIMRUTH
Avatar
Send message
Joined: 24 Feb 05
Posts: 12
Credit: 1,886
RAC: 0
Message 5882 - Posted: 27 Feb 2005, 12:57:54 UTC

I guess the first thing to do is get a better cooler.. or get me one of them watercooling kits.. maybe it will hold at 1900mhz..

Right now I`m using the boxed sempron, so not really suitable for high oc`s
____________

Profile Santas little helper
Send message
Joined: 11 Feb 05
Posts: 37
Credit: 4,223,294
RAC: 4,450
Message 5901 - Posted: 27 Feb 2005, 14:58:47 UTC
Last modified: 27 Feb 2005, 15:56:23 UTC

HW: P4 2,67@2,7 / 512MB@135MHz / 1384 Whetstones / 4189 Dhrystones
SW: WinXP Pro SP2 / BOINC 4.19 / Einstein 4.79

WU duration: something around 8,75h
____________
Greetings, Santas little helper

Profile Thierry Van Driessche
Avatar
Send message
Joined: 9 Feb 05
Posts: 210
Credit: 229,929
RAC: 0
Message 5905 - Posted: 27 Feb 2005, 15:20:14 UTC
Last modified: 27 Feb 2005, 15:23:23 UTC

P4 HT 2.4@2.88GHz, bus speed 960MHz, 512MB RAM@160MHz. CPU temp 44°C/111°F with room temp around the 20°C/68°F.

Running on Win XP Pro SP2. Boinc v4.19 and Einstein 4.79.

Crunching 2 WU's at the same time. CPU time per WU between 11:10 and 12:20h.

Profile Oliver
Avatar
Send message
Joined: 24 Feb 05
Posts: 56
Credit: 129,576
RAC: 0
Message 5906 - Posted: 27 Feb 2005, 15:31:24 UTC

Profile Dominique
Avatar
Send message
Joined: 19 Feb 05
Posts: 67
Credit: 16,797
RAC: 0
Message 5911 - Posted: 27 Feb 2005, 15:58:10 UTC
Last modified: 27 Feb 2005, 15:58:22 UTC

AMD Barton XP2400-M @2.1ghz 512k L2
512mb PC3200 ram

E@H: 6 hr
S@H: 2.5-2.75 hr
P@H: 55 min


____________

*I still know CRAP when I see it.

NIMRUTH
Avatar
Send message
Joined: 24 Feb 05
Posts: 12
Credit: 1,886
RAC: 0
Message 5915 - Posted: 27 Feb 2005, 16:03:22 UTC

as I suspected... even the cheapest AMD, the sempron 2200+ I have is kicking the butts off most Intel P4 systems.. now how funny is that ?

Are the semprons so good or are the P4`s so bad ?

now that I oc`d the 2200+ to 1800mhz, my time is 6:34 hrs per WU for einstein and 3.75 for S@H
____________

Profile Jordan Wilberding
Send message
Joined: 19 Feb 05
Posts: 162
Credit: 715,454
RAC: 0
Message 5918 - Posted: 27 Feb 2005, 16:09:05 UTC - in response to Message 5915.

> as I suspected... even the cheapest AMD, the sempron 2200+ I have is kicking
> the butts off most Intel P4 systems.. now how funny is that ?
>
> Are the semprons so good or are the P4`s so bad ?
>
> now that I oc`d the 2200+ to 1800mhz, my time is 6:34 hrs per WU for einstein
> and 3.75 for S@H
>

I was always under the assumption that sempron's were worse than an Athlon XP. But my Athlon XP 2400+ takes almost 12 hours for a WU.
____________
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell

Profile Thierry Van Driessche
Avatar
Send message
Joined: 9 Feb 05
Posts: 210
Credit: 229,929
RAC: 0
Message 5922 - Posted: 27 Feb 2005, 16:18:37 UTC - in response to Message 5915.

> Are the semprons so good or are the P4`s so bad ?
>
> now that I oc`d the 2200+ to 1800mhz, my time is 6:34 hrs per WU for einstein
> and 3.75 for S@H

Well, don't underestimate the P4 HT:

3.75h. for S@H means you are doing some 6.4WU's a day.

With my P4 HT 2.4@2.88GHz, I'm doing real time some 10 to 11 WU's a day ;)

____________
Greetings from Belgium
Thierry

Profile ErichZann
Avatar
Send message
Joined: 11 Feb 05
Posts: 121
Credit: 81,582
RAC: 0
Message 5928 - Posted: 27 Feb 2005, 16:39:07 UTC - in response to Message 5918.


> >
>
> I was always under the assumption that sempron's were worse than an Athlon XP.
> But my Athlon XP 2400+ takes almost 12 hours for a WU.
>

Yes, Athlon XP's are normaly faster... and as i said i need 7 hours for a e@h WU with my 2200+... something doesnt seem to be alright with yours
____________

Profile Jordan Wilberding
Send message
Joined: 19 Feb 05
Posts: 162
Credit: 715,454
RAC: 0
Message 5930 - Posted: 27 Feb 2005, 16:42:40 UTC - in response to Message 5928.

>
> > >
> >
> > I was always under the assumption that sempron's were worse than an
> Athlon XP.
> > But my Athlon XP 2400+ takes almost 12 hours for a WU.
> >
>
> Yes, Athlon XP's are normaly faster... and as i said i need 7 hours for a e@h
> WU with my 2200+... something doesnt seem to be alright with yours
>

Hmmm, well I have several 2400+ machines, and they are all around this speed. I wonder what it could be?
____________
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell

Seti-Cruncher
Send message
Joined: 9 Feb 05
Posts: 70
Credit: 7,114
RAC: 0
Message 5931 - Posted: 27 Feb 2005, 16:48:04 UTC - in response to Message 5918.

> I was always under the assumption that sempron's were worse than an Athlon XP.
> But my Athlon XP 2400+ takes almost 12 hours for a WU.

That's because you are running Linux. The Einstein client performs very badly in Linux. In SETI, there is little, if any, difference between Windows and Linux.



____________
Be lucky,

Neil

Profile BlackAdder
Avatar
Send message
Joined: 9 Feb 05
Posts: 2
Credit: 561,686
RAC: 0
Message 5933 - Posted: 27 Feb 2005, 16:54:24 UTC

I'm running a XP3200, Asus A7N8X delux MB, 1.5 gig Corsair DDR 400 ram, nothing overclocked. My Einstein times are 5:50 to 6:00 hours, seti Boinc 2:30, Mfold 1.25 55 minutes average.
____________

KWSN_Dagger
Send message
Joined: 22 Jan 05
Posts: 17
Credit: 5,180
RAC: 0
Message 5945 - Posted: 27 Feb 2005, 18:12:21 UTC

AMD64 3200+ with Win ME: SETI 2.5 to 3 hrs, Einstein 4 to 5 hrs, Predictor 40 to 50 mins. LHC 4 seconds to 2 hrs.
____________

Profile Jordan Wilberding
Send message
Joined: 19 Feb 05
Posts: 162
Credit: 715,454
RAC: 0
Message 5946 - Posted: 27 Feb 2005, 18:13:46 UTC - in response to Message 5931.

> > I was always under the assumption that sempron's were worse than an
> Athlon XP.
> > But my Athlon XP 2400+ takes almost 12 hours for a WU.
>
> That's because you are running Linux. The Einstein client performs very badly
> in Linux. In SETI, there is little, if any, difference between Windows and
> Linux.


Why is this? You would think in Linux they could make it even better than for Windows.
____________
such things just should not be writ so please destroy this if you wish to live 'tis better in ignorance to dwell than to go screaming into the abyss worse than hell

NIMRUTH
Avatar
Send message
Joined: 24 Feb 05
Posts: 12
Credit: 1,886
RAC: 0
Message 5956 - Posted: 27 Feb 2005, 19:13:12 UTC

okay...

after several crashes I have settled for 1710Mhz with a 190 FSB and a Vcore of 1.67 Volts..

I cranked it up to 1900 Mhz for a few minutes at 210fsb.. it crashed.. at 60C it shut down

I stand corrected for the Einstein WU time.. it takes me 7hrs36mins to complete.. But that`s at 1710mhz, once my watercooler is hooked up, I can run stable at 1850mhz or so, that would amount to aprox. 30mins speedup..


Basicly, all indications are that AMD beats the Intel P4 at any test..

I`ll never buy intel again..
____________

Seti-Cruncher
Send message
Joined: 9 Feb 05
Posts: 70
Credit: 7,114
RAC: 0
Message 5980 - Posted: 27 Feb 2005, 20:48:20 UTC - in response to Message 5946.

> Why is this? You would think in Linux they could make it even better than for
> Windows.

Compiling for Linux is a lot more complicated than compiling for Windows because of the far more diverse range of platforms the compiler can target for. They simply haven't found the "right" compiler optimisations yet. Actually, it's even more complicated than that because the libraries linked (either dynamically or statically) need to be optimised too to have any real effect.

It's a shame the source code has not been released - otherwise some of the people out here with more time to spare may have cracked it by now. ;)


____________
Be lucky,

Neil

1 · 2 · 3 · 4 . . . 10 · Next

Message boards : Cruncher's Corner : REPORT WU-SPEEDS HERE !


Home · Your account · Message boards

This material is based upon work supported by the National Science Foundation (NSF) under Grants PHY-1104902, PHY-1104617 and PHY-1105572 and by the Max Planck Gesellschaft (MPG). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the investigators and do not necessarily reflect the views of the NSF or the MPG.

Copyright © 2016 Bruce Allen