S5R2

Annika
Annika
Joined: 8 Aug 06
Posts: 720
Credit: 494410
RAC: 0

Mine is a 666 MHz P3 with 256

Mine is a 666 MHz P3 with 256 MB of memory (yeah!), so, not one of the really fast ones but as you said, it should be able to manage. Btw, the largest WUs I know are about 530 credits worth and run for about 28 hours on an AMD 64 3500+ (Linux, so, no AMD penalty) and I keep getting those.

Bikeman (Heinz-Bernd Eggenstein)
Bikeman (Heinz-...
Moderator
Joined: 28 Aug 06
Posts: 3522
Credit: 728657444
RAC: 1176800

RE: Mine is a 666 MHz P3

Message 62463 in response to message 62462

Quote:
Mine is a 666 MHz P3 with 256 MB of memory (yeah!), so, not one of the really fast ones but as you said, it should be able to manage. Btw, the largest WUs I know are about 530 credits worth and run for about 28 hours on an AMD 64 3500+ (Linux, so, no AMD penalty) and I keep getting those.

E@H was so kind to assign one of those monster WUs to my PIII 866 Coppermine, estimated CPU time (usually a very precise estimation) is 82h (!) compared to the usual 51h WUs my Coppermine gets (for ca. 300 credits). That host is running 24/7, so I don't mind and have no deadline problem. I hope the server-side logic is so smart to be a bit careful about assigning these huge WUs to older and less active computers.

@archae86: I guess you pay a big MS WIN 98 penalty here, my Coppermine (under Linux) is about 33% faster but has lower clock frequency and less cache (is yours a PIII Xeon??)

CU

BRM

archae86
archae86
Joined: 6 Dec 05
Posts: 3157
Credit: 7224934931
RAC: 1043477

RE: ... my PIII 866

Message 62464 in response to message 62463

Quote:

... my PIII 866 Coppermine, estimated CPU time (usually a very precise estimation) is 82h (!) compared to the usual 51h WUs my Coppermine gets (for ca. 300 credits).
...

@archae86: I guess you pay a big MS WIN 98 penalty here, my Coppermine (under Linux) is about 33% faster but has lower clock frequency and less cache (is yours a PIII Xeon??)

Mine is just a Dell Dimension consumer box, definitely not a Xeon. Maybe this is a case where the Linux version of the Einstein code is much more efficient. I don't think Win98 itself is burning anything like a third of the CPU time, so, much as I enjoy bashing Microsoft performance, I doubt that is an issue here.

You've hidden your computers, so I can't comment further.

From a power efficency point of view, the Green thing for both of us would be to turn off our Coppermine machines and run modern power-efficient processors instead.

Oh, come to think of it, I just did that--turned off a 733 MHz Coppermine Win98 machine and replaced it with a 3 GHz Conroe WinXP machine. What a difference!

[edit to correct two spelling errors immediately after initial post]

Annika
Annika
Joined: 8 Aug 06
Posts: 720
Credit: 494410
RAC: 0

Hey, that's a new approach...

Hey, that's a new approach... I've been annoyed about my Dad not wanting to retire that old box (both because I have to do "support", which sucks on such a slow machine running XP, and because he lets me crunch on his box and a better PC would do more for the science and get more credit), so maybe it's worth trying telling him that it's bad for the environment :-D

RandyC
RandyC
Joined: 18 Jan 05
Posts: 6608
Credit: 111139797
RAC: 0

Well, the XP2200+ came in and

Message 62466 in response to message 62403

Well, the XP2200+ came in and I swapped the XP1700 to another system, replacing the Athlon 1333 CPU.

On the new MB, BOINC says:

Quote:
5/3/2007 7:15:39 PM Processor: 1 AuthenticAMD AMD Athlon(tm) Processor [x86 Family 6 Model 8 Stepping 1] [fpu tsc sse 3dnow mmx]

Per BOINC, it DOES has SSE on it. Something in the old MB must keep the system from seeing SSE (the SETI SSE science app would blow up when I tried to use it).
[edit]BTW, CPUz sees it on the new MB as well, but didn't on the old one.[/edit]

Quote:
Quote:
Quote:
S5R2 using SSE, S5R1 wasn't, as per Akosf's post?

S5R1 had FPU and SSE optimised subroutines too.
S5R2 doesn't use SSE code at moment.

Quote:
Per CPUz, that particular CPU is XP1700+ Thoroughbred with MMX(+), and 3DNOW(+). No SSE on it.

I had four Athlon XP 1700+ CPUs, all of them knew SSE.
I went over the AMD Tech Docs to check myself.
All of the Athlon XP CPUs know SSE instructions.

AMD Athlon XP Tech docs:
24309 AMD Athlonâ„¢ XP Processor Model 6 Data Sheet (Palomino)
25175 AMD Athlonâ„¢ XP Processor Model 8 Data Sheet (Thoroughbred)
26237 AMD Athlonâ„¢ XP Processor Model 10 Data Sheet (Barton)

Well, this particular one doesn't. My XP 1600+ system (before it caught on fire) would run rings around it...IT definitely had, and used, the SSE instructions! I've always been disappointed in that particular chip. It's due to be replaced with an XP 2200+ ($46) this coming week. Just about the max that MB will take.


Seti Classic Final Total: 11446 WU.

DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1364
Credit: 3562358667
RAC: 0

RE: Per BOINC, it DOES has

Message 62467 in response to message 62466

Quote:

Per BOINC, it DOES has SSE on it. Something in the old MB must keep the system from seeing SSE (the SETI SSE science app would blow up when I tried to use it).
[edit]BTW, CPUz sees it on the new MB as well, but didn't on the old one.[/edit]

At a guess it needed a bios flash. IF the bios predated the XP chips it wouldn't know how to turn the feature on.

RandyC
RandyC
Joined: 18 Jan 05
Posts: 6608
Credit: 111139797
RAC: 0

RE: RE: Per BOINC, it

Message 62468 in response to message 62467

Quote:
Quote:

Per BOINC, it DOES has SSE on it. Something in the old MB must keep the system from seeing SSE (the SETI SSE science app would blow up when I tried to use it).
[edit]BTW, CPUz sees it on the new MB as well, but didn't on the old one.[/edit]

At a guess it needed a bios flash. IF the bios predated the XP chips it wouldn't know how to turn the feature on.

That's a possibility I'll have to check out. The XP2200+ doesn't show SSE either using CPUz or BOINC.

MB is ECS K7S5A

Seti Classic Final Total: 11446 WU.

Arion
Arion
Joined: 20 Mar 05
Posts: 147
Credit: 1626747
RAC: 0

RE: That's a possibility

Message 62469 in response to message 62468

Quote:

That's a possibility I'll have to check out. The XP2200+ doesn't show SSE either using CPUz or BOINC.

MB is ECS K7S5A

I'm using a Soyo KT880 MB with an AMD 2200+ CPU and CPUz version 1.39 shows it has MMX (+), 3DNow! (+), SSE

WU's are running for 18+ hours and receiving 169 - 170 credits.

[edited to correct credits]

Bikeman (Heinz-Bernd Eggenstein)
Bikeman (Heinz-...
Moderator
Joined: 28 Aug 06
Posts: 3522
Credit: 728657444
RAC: 1176800

RE: RE: ... my PIII 866

Message 62470 in response to message 62464

Quote:
Quote:

... my PIII 866 Coppermine, estimated CPU time (usually a very precise estimation) is 82h (!) compared to the usual 51h WUs my Coppermine gets (for ca. 300 credits).
...

@archae86: I guess you pay a big MS WIN 98 penalty here, my Coppermine (under Linux) is about 33% faster but has lower clock frequency and less cache (is yours a PIII Xeon??)

Mine is just a Dell Dimension consumer box, definitely not a Xeon. Maybe this is a case where the Linux version of the Einstein code is much more efficient. I don't think Win98 itself is burning anything like a third of the CPU time, so, much as I enjoy bashing Microsoft performance, I doubt that is an issue here.

You've hidden your computers, so I can't comment further.

From a power efficency point of view, the Green thing for both of us would be to turn off our Coppermine machines and run modern power-efficient processors instead.

Oh, come to think of it, I just did that--turned off a 733 MHz Coppermine Win98 machine and replaced it with a 3 GHz Conroe WinXP machine. What a difference!

[edit to correct two spelling errors immediately after initial post]

Upps, sorry, I've now "unhidden" my computers. This is the one I was talking about:

http://einsteinathome.org/host/732766

I didn't know there was a regular P III Coppermine with 1 MB cache, so I suspected your's is a Xeon. Maybe the cache figure is wrong.

As for the green thing: Yup, I just did that as well, an old 800 Mhz SSE-less Athlon was retired, and today I'll power up a new Mac mini :-). I think I will use that one at least 50% for CPDN, tho.

CU

BRM

Annika
Annika
Joined: 8 Aug 06
Posts: 720
Credit: 494410
RAC: 0

Really sounds much to me for

Really sounds much to me for a Coppermine. Ours only has 256 KB of cache (but then, it was a discounter box... for those of you who also come from Germany, the term "Medion" should ring a bell...) and I thought they only came like that. Another lesson learned ;-) that's one reason why I like these message boards.
Bikeman, gratz to your new box. I have no idea about Apples, but it sounds like a nice little machine. And so fitting to use it for CPDN after going the "green" route about the old box ;-) I also used to do some work for that project, it's quite nice to never run out of work... but I hope you have a few percent left for Einstein ;-)
I won't be getting new boxes any time soon (no wonder, just bought two over the last year) but maybe someone in my family will, and that way they would join my farm aswell. But most of those don't do Einstein atm. My brother's notebook has a Dothan CPU (Pent M, he bought it just so early that he missed out on dual cores, but it's a nice little machine nevertheless) with a really large cache (2 MB), so it's just right for some projects I couldn't do too well otherwise. As he doesn't like SETI, I attached the box to MalariaControl and it does really well there. Those two projects seem to prefer similar architectures. The Coppermine was offline for about 10 days, forcing me to abort a WU (sorry to my wingman) because we had to order an ethernet card first; it used to connect via USB but the new router hardware doesn't support that. That one is doing Einstein as well as a bit of PrimeGrid. Hope the WUs won't be too large, but it'll manage. Oh yeah, my mother is using my old Banias notebook now (says it's just fine for Office and stuff, which is probably true) but that will be PrimeGrid only, as soon as Microsoft fixes those annoying DLL errors which won't let me install BOINC. My own two boxes, the Athlon and the C1D, are doing 100% Einstein atm. Normally I'd also do DRTG on the Athlon. As for the notebook, I considered reattaching to SETI but they don't seem to have as much work as the mail said, so, not sure yet. I'm also one of the people waiting for Orbit@home to finally have work, but I have no idea yet which box(es) I'm going to run it on.
So much for my little "veggie patch", as someone called really small farms. I do hope I'll be able to convince some people at Uni to let me attach some boxes there. No idea what they use, I think many of them are Netbursts but they're online a lot...

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.