• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

why are celerons slow at folding?

G

grethor

Limp Gawd
Joined
Mar 1, 2004
Messages
188
I installed FAH4Console -forceSSE -advmethods on my girlfriends box ( celeron 900 MHz with Win98 and 128 MB ram ) and i must say that it's dog slow. It's being outpaced by my P3 450 MHz box running linux and with 384 MB ram. What major architectural differences are there between the PIII and the Celeron?
 
well there are multiple celerons IIRC, a pIII based one and a P4 based one, but in any case, celerons are basically dog slow at everything. Their are multiple reasons...one, I think they have less cache, at least they used to. But the main difference, again if IIRC, is that they execute fewer instructions per clock cycle than the chips they were derived from, so they do less work per cycle. This obviously makes them slower. On a related note, this is how AMD's can have a PR matching that of a higher clocked pentium chip, because the AMDs execute MORE per cycle.

Or something like that? :)
 
It's a celeron 900 so I assume that it's P3 based. I just find it interesting how the celeron is being beat by a P3 that has half the clockspeed.
 
Different proteins take different amounts of time per frame to finish. The Celeron might be crunching a protein that is suppossed to take longer. Check which protein each is working on and then you can get an idea of the size of each one here.

But yeah, Celerons are generally slow. Maybe not slower than a cpu half it's speed but still pretty slow.
 
large difference in RAM as well, plus are both boxen dedicated folders or are they being used at the same time?
 
i'd say the 1/2 the cache difference (IIRC) and the 1/3 of the RAM are hurting her box pretty bad...you should upgrade it for her :D
 
my celly 800 that is clocked to 1.03ghz with 384mb or ram right now is still on its first WU thingy, and my 2.4ghz amd system in my sig is on #9... i started them at the same time about a week ago...

so does that mean my athlon system is 9 times faster??
 
are they folding the same protein? is it a tinker or a gromacs core?

amds tear through tinkers
 
I was running gromacs on both of them. The 450 can blow through a gromacs 100 frame ( the one that folds fast ) in about a day while the celly only gets through about a third of it in about a day. I dont think that lack of ram is an issue because i looked at the FAH32.exe memory consumption on my athlon xp 2500, winXP box and it was only using about 11 megs.
 
scottatwittenberg: Welcome to the team!! let the addiction begin :D


 
thanks...

also.. all of the WU's (what does that mean anyways) are either 400 or 1000 something or others... actually my xp system is doing a 500 now... but this damn celeron is still on its first one...
 
WU = work unit = a protein.

If you say you've done 300 WUs, your boxen combined have completely finished 300 proteins that stanford has sent you
 
My A-XP 2500+ is literally 6 times faster at folding than my P4 2.4GHz.... Wonder why that is...

5WU's on P4 = 45 minutes
5WU's on AXP = 7 minutes
 
Jeez guys - it seems like you all are standing around discussing George W. Bush's intelligence. You've got no DATA. Try and give some performance numbers that mean something. I've machines working on Work Units (proteins) which vary in credit from 30 to 242 points.
For some reason the 242 pointers take a BIT LONGER!!!.
Monitor the performance with EMIII and tell us in credit units per day per Ghz, or Points per Day per Ghz (PDG). That is understandable.
 
AMD has always had an edge in math, thus it's faster at folding. I have one Intel 3.2 P-4 and an AMD at 2.2 is simply faster. Turned off my one gig celery a year ago, wasn't worth the power it took.

Fold on

I'll let the "Bush" comment pass for the moment ;)

BillR
 
BillR said:
AMD has always had an edge in math, thus it's faster at folding. I have one Intel 3.2 P-4 and an AMD at 2.2 is simply faster. Turned off my one gig celery a year ago, wasn't worth the power it took.

Fold on

I'll let the "Bush" comment pass for the moment ;)

BillR
Click to expand...
bah, yay for the Bush comment :p

Hmm, does that mean that i should get an XP-M to OC ? I was planning on getting a really OCable system (i was, up till now, thinking p4 2.4c) and WCing it w/ the Koolance Exos Al WCing system (since this'll be my first foray into WCing) for folding. Should i go the AMD route instead?
 
You must log in or register to reply here.
Back
Top