• 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.

3.06Ghz = 52sec SuperPI??!!?

C

Citizen _insane

Gawd
Joined
May 2, 2006
Messages
565
I'm running dual Xeon Prestonias @ 3.06Ghz (stock speed, 133x23) and I'm getting 52 second SuperPi scores. That seems slow to me. Any ideas what could be bottlenecking my system? I know that dual processors does nothing for SuperPi, but for a 3ghz, 52 seems slow. I've been having random freezes recently too, I'm starting to think my motherboard is dying.
 
My 3.2ghz p4 gets 47 secs, so thats normal. Get a conroe if you want super fast super pies.
 
forcefed said:
My 3.2ghz p4 gets 47 secs, so thats normal. Get a conroe if you want super fast super pies.
Click to expand...
super pies
meet pies
veggie pies
apple pies

they all the same m8ty
 
lol, mmm muff pie! ;)



that is normal, you have an old processor with a slow FSB and technology compared to the conroe;s and such today getting sub 20 second PI times.
 
i dunno, my 3.0ghz P4 scored around 38 seconds, but it's LGA775. I'd probably format the hard drive. my friend's 3.2E got around 57 second 1Mb pi calcs and after a format he pulled about 20 seconds faster. maybe the hard drive is stuffed with crap, lol.
 
StealthyFish said:
i dunno, my 3.0ghz P4 scored around 38 seconds, but it's LGA775. I'd probably format the hard drive. my friend's 3.2E got around 57 second 1Mb pi calcs and after a format he pulled about 20 seconds faster. maybe the hard drive is stuffed with crap, lol.
Click to expand...
I didnt think the HDD had ANYTHING to do with a PI test :confused: If I give this some thought, after I pulled my 24 second 1MB PI run on my 3GHz FX57, if I formatted my HDD I could have pulled a 4 second PI run, UMMMM ;)
 
52 seconds is probably about right. Superpi really just tests memory bandwidth and latency, neither of which a 533fsb smp xeon setup is going to do exceptionally well on. That's why I don't see superpi as a very good cpu benchmark...it relies more on the memory subsystem.
 
Yes, superpi is big on memory bandwidth. Lower your memory timings and i gurantee your superpi time will go down.

I was playing with a celeron D 3.06ghz today and it did it in 52sec. Talk about slow. Even at 4.5ghz it only came in at 41sec. Sloooow.
 
newls1 said:
I didnt think the HDD had ANYTHING to do with a PI test :confused: If I give this some thought, after I pulled my 24 second 1MB PI run on my 3GHz FX57, if I formatted my HDD I could have pulled a 4 second PI run, UMMMM ;)
Click to expand...


super pi is logarithmic, not linear decrease. What you just said made no sense becuase it's nowhere near directly proportional. A lot of shit decreased my friend's performance, and his calculations did improve. I don't get why a 3.2Ghz processor calculates slower than my 3.0ghz. Maybe socket change? same design, i've got a prescott, he's got a prescott. only difference i saw was the temps, where his was higher (as expected with S478). and considering that my friend's hard drive wasn't formatted for about 2 years and he downloaded a lot of random shit like new cursors and other useless stuff, it really messed his performance. You want to use my P4 willamette which wasn't formatted in 5 years? There's a noticable performance loss.
 
StealthyFish said:
super pi is logarithmic, not linear decrease. What you just said made no sense becuase it's nowhere near directly proportional. A lot of shit decreased my friend's performance, and his calculations did improve. I don't get why a 3.2Ghz processor calculates slower than my 3.0ghz. Maybe socket change? same design, i've got a prescott, he's got a prescott. only difference i saw was the temps, where his was higher (as expected with S478). and considering that my friend's hard drive wasn't formatted for about 2 years and he downloaded a lot of random shit like new cursors and other useless stuff, it really messed his performance. You want to use my P4 willamette which wasn't formatted in 5 years? There's a noticable performance loss.
Click to expand...
Everything I said in my post, was a play on words based on what you said ;) . So it must make sense, since you said it :eek: Im formatting my HDD now, I'm hoping for a 10sec PI run.............
 
formatting a HD will have very little effect, unless your PC is completely full of bloatware, spyware, etc.... and that just means you had a sloppy computer and a fresh install will run everything a little faster b/c you don't have the bloatware yet. But I SERIOUSLY doubt anything close to a 20 second difference.

BTW, the same probably could have been done by being smart and shutting down extra services & start up items, using msconfig & admin tools.

back to the OP, 52 seconds is about right for that CPU.
 
He says hes running dual Xeon 3.06's. I think he should be running alot faster than that. My 2.4C@3.4 is does 1M in around 40s.
 
Slow memory. Going from value ddr400 with cl2.5 to ddr2 667 4-4-4-12 moved my 805d from 58-59s to 54-55s without OC.
 
the 533 memory bus is what is slowing it down.
my 2.6 northwood at stock clocks can do 1M in 50 seconds because of the faster 800 FSB and dual channel.
 
acascianelli said:
He says hes running dual Xeon 3.06's. I think he should be running alot faster than that. My 2.4C@3.4 is does 1M in around 40s.
Click to expand...

wrong

SuperPi isn't multi-threaded so dual CPU's won't help.

Your 2.4C oc'd should be much faster b/c you're on an 1133 FSB while he's on a 533 FSB (2x faster) and its a little over 10% faster CPU speed.

So the combination of much faster bus, much higher bandwidth/lower latency memory and faster CPU is why you're at 40 & he's at 52.
 
newls1 said:
Everything I said in my post, was a play on words based on what you said ;) . So it must make sense, since you said it :eek: Im formatting my HDD now, I'm hoping for a 10sec PI run.............
Click to expand...

I'm sorry, are you retarded or something? Your X-2 couldn't hit anywhere near that if you dreamed, and his Pi score was horrible after his hard drive had enough crap to actually hinder his performance. You don't make any sense. Don't post if you don't get it and don't know jack.
 
StealthyFish said:
I'm sorry, are you retarded or something? Your X-2 couldn't hit anywhere near that if you dreamed, and his Pi score was horrible after his hard drive had enough crap to actually hinder his performance. You don't make any sense. Don't post if you don't get it and don't know jack.
Click to expand...

dude he's just pulling your chain...like he said, he was just messing with your words.
 
newls1 said:
I didnt think the HDD had ANYTHING to do with a PI test :confused: If I give this some thought, after I pulled my 24 second 1MB PI run on my 3GHz FX57, if I formatted my HDD I could have pulled a 4 second PI run, UMMMM ;)
Click to expand...


The HDD doesn't have anything to do with it. But steatlhy is talking about background programs and stuff running off that hard drive which may be clogging up the system. But this time seems ok for a Xeon, a little slow.
 
StealthyFish said:
I'm sorry, are you retarded or something? Your X-2 couldn't hit anywhere near that if you dreamed, and his Pi score was horrible after his hard drive had enough crap to actually hinder his performance. You don't make any sense. Don't post if you don't get it and don't know jack.
Click to expand...
who the hell said anything about an X2 chip :confused:
 
Crazy my 2.8ghz Northwood clocked @3.0 with 1gig DDR at 179mhz 2.5-3-2-5, just did 49sec 1mb.
I thought for sure your system would beat mine. And I even had stuff running.
 
I overclocked my proc's to 3.3Ghz and saw an increase to 47sec. However, I then increased it to 3.45Ghz and ran only 46sec. The system isn't fully stable at 3.45, it fails Prime95, but passes SuperPi 32M and memtest. Anyway, I'm running ths system at 3.3 daily now and its perfecly stable.
 
iLLuSioN said:
Crazy my 2.8ghz Northwood clocked @3.0 with 1gig DDR at 179mhz 2.5-3-2-5, just did 49sec 1mb.
I thought for sure your system would beat mine. And I even had stuff running.
Click to expand...

Is that a 2.8 with 800FSB? overclocked?

even if it was a 2.8 @ 533FSB, you overclocked it

so both have a faster FSB than his 3.06 @ 533 stock processor.
 
so i am still a bit confused is 52sec "normal" for a dual Xeon setup? If it is more of a memory interface it wouldnt matter if was dual cpu or not because it seems as fast as a single cpu

On a side note I ran my roomates hp pc with 2.5 northwood and a 1m run netted something like 2+minutes I think it was more like 4min just dont remember . it uses pc2100 512mb ram 533bus? I was thinking of reformatting the hdd
 
I get 57 seconds on my northwood at 3.34. The only reason I can think that my score isn't as fast as some of the other northwoods in the vicinity of my clock speed is that I have a 1 gig PC2100 memory coupled with a 512 mb PC3200 memory. Since I'm overclocking, I have to run the memory with a 3/4 divider. If I were at 1/1, My memory would be running at 320mhz. Instead I'm stuck with 240mhz.

Hard to believe that can make a 10 second difference in super pi score though. Maybe I'll throw it back to stock mhz and 1/1 and test through.
 
Nope... 1minute 1 second on bone stock 2.8

How the hell are people getting into the 40s with the northwood? Is my system just jacked?

Does anyone have a link to a database here to shows users times and their system components?
 
that seems slow to me...My 3.2Ghz Northwood overclocked to 3.85ish can calculate SuperPI in about 30 seconds.

My memory is clocked at a 1:1
 
Is that a 800mhz bus or 533? Maybe that's the difference here.
 
my Pentium D 920 (OC'd to 3.9Gz) does it in 33.23 seconds.

FSB + memory speed & timings + CPU speed all make a difference.
 
Flak Pyro said:
My 3.06 533FSB cpu scrores around the same as yours last time i ran super pi.
Click to expand...

Thanks for the reassurance that something is not overly wrong with my computer.

What is the deal with super pi anyway? I've heard of it being thrown around in forums for ages but didn't actually go out to test my system until all the hoopla recently of an overclocked Conroe doing the 1M superpi in 10 seconds. That's when I finally decided to test my own system.

Does it translate to real world performance? I guess for me the only real world performance would have to relate to games. And I can't see how someone else with a northwood running the same mhz as me, besting me by 10 seconds on the superpi, is going to perform at that same percentage better than me in terms of frames per second. Am I correct in this?
 
well videocard is FAR more important for gaming. So if you had the same videocard, you might get a few less fps. Nothing big, though. Super PI is a straight CPU & memory test. Conroe does very well b/c of its large 4Mb L2 cache, plus its P3 like structure(very efficient).
 
the conroe cpu derived from both the Pentium M line and the Prescott P4 line. all they did was take the best from each line and combined them together to get what is now known as conroe
 
You must log in or register to reply here.
Back
Top