Super Pi Scores, Post yours here.

Sojourn · Post by **Sojourn** » 21 Oct 2005, 10:26

dammod wrote:Because I wanted to. Was bored. Oh....and because I can

Since I also used that excuse before when I had nothing better to come up with, I guess it's OK.

s

dammod · Post by **dammod** » 21 Oct 2005, 11:41

Sojourn wrote:
dammod wrote:Because I wanted to. Was bored. Oh....and because I can
Since I also used that excuse before when I had nothing better to come up with, I guess it's OK.

s

Post by **Thrall** » 21 Oct 2005, 16:37

36 secs here on an AMD 4000 (stock)

Abominog · Post by **Abominog** » 21 Oct 2005, 16:53

Haha, my workshop PC, Celeron 1100, 256megoram

512K 3min55s

Onnodig!!

AlphA · Post by **AlphA** » 24 Oct 2005, 23:37

I get 53 secs on a stock P4 3.0Ghz

Basipooh · Post by **Basipooh** » 24 Oct 2005, 23:40

AlphA wrote:I get 53 secs on a stock P4 3.0Ghz

that cant b right surely........?
u did the 1M test???
i got 39s with my p4 at 3.0ghz stock and 37 when it was at 3.6GHz.

AlphA · Post by **AlphA** » 25 Oct 2005, 00:07

Basipooh wrote:
AlphA wrote:I get 53 secs on a stock P4 3.0Ghz
that cant b right surely........?
u did the 1M test???
i got 39s with my p4 at 3.0ghz stock and 37 when it was at 3.6GHz.

Yup, unfortunately, that is right....

I got 53 seconds...just clocked it up to 3.3Ghz (220*15) and now I get 48.

What could be my problem??

AlphA · Post by **AlphA** » 25 Oct 2005, 01:06

I think I know what my problem was.....I had HT enabled in the BIOS and it wasn't utilizing the full CPU to do the test. I disabled it , and now I got 40 secs for a 3.0Ghz @ 3.3Ghz..

Been looking at some scores, thats not too bad considering I have Kingston valueram (3-3-3)

Basipooh · Post by **Basipooh** » 25 Oct 2005, 07:04

i did my test with HT enabled

g3ck0 · Post by **g3ck0** » 25 Oct 2005, 09:21

29.980 on A64 3700+ (overclocked on stock heatsink) Cant wait to "seat" the Vapo though =)

http://www.overclocking.co.za/index.htm ... illisecond

AlphA · Post by **AlphA** » 25 Oct 2005, 10:48

basipooh - then do you have any idea why mine was slower with HT enabled??

Basipooh · Post by **Basipooh** » 25 Oct 2005, 12:52

No idea, The only thing i can think of (and its a little far fetched)
is:

The new prescott core has 2mb l2 cache (compared to mine with 1mb)
Tho it is a bigger cache than the old prescott, it is a slower cache the old prescott.

What HT does is divide the cpu into to a floating point proccesor and an integer processor AND it shares the cache between the two virtual cores
leaving 1mb of SLOW cache per core.

Since super pi is mainly floating point calculations and does not require a lot of memory, a smaller fast cache would be prefrable to a large slow cache.
Hence the reason why my pressie with 512k cache for the FLOATING point processor (and the unused 512k on the integer proccesor) performed the Pi-calculation quicker.

When u disable HT it allows the full 2mbs of cache to be used therefore making up for the speed loss by having a huge cache (therefore giving u a faster time)

*
I just sucked this out my thumb, can someone let me know if this has any truth to it coz it makes sense in a wierd way.

Post by **Anakha56** » 25 Oct 2005, 15:01

does any1 know of an http d/l for this proggi? i can d/l from ftp here @ work...

*edit* found 1...

Post by **Anakha56** » 25 Oct 2005, 15:19

alpha dont feel bad my work proc, see thumbnail also does it in 45sec...

am going to slightly overclock this guy to see what it can do...

AlphA · Post by **AlphA** » 25 Oct 2005, 15:56

hehe - thanks for the moral support Anakha

Once I get my Vantec fans, I'll be able to squeeze a bit more out of it...

Yah, Basipooh, maybe the proc does'nt make full use of the CPU's abilities...

AlphA · Post by **AlphA** » 26 Oct 2005, 22:27

After clocking to 3.6Ghz .... I now got 37 secs

VoodooProphetII · Post by **VoodooProphetII** » 27 Oct 2005, 06:31

Well done Alpha.......well done

Jasonhk · Post by **Jasonhk** » 27 Oct 2005, 07:01

Anakha56 wrote:does any1 know of an http d/l for this proggi? i can d/l from ftp here @ work...

*edit* found 1...

Well here is another one:

http://196.35.73.164/sa3doc/downloads/s ... od_1.4.zip

Locally Hosted as well

jPm · Post by **jPm** » 27 Oct 2005, 08:57

I get 37 seconds on the 1meg test without overclocking at all with HT or without HT

AlphA · Post by **AlphA** » 27 Oct 2005, 09:12

VoodooProphetII wrote:Well done Alpha.......well done

Thanks a lot man! Appreciate it.

jpm --> How come it makes no difference to you whether you have HT enabled or not?? Which OS are you running?

jPm · Post by **jPm** » 27 Oct 2005, 09:43

not sure why, i tryed it with both and get the same result,im using windows xp x64 and igot a em64t processor

any ideas why it doesnt make a difference

would i get a better pi on the 32bit xp pro?

AlphA · Post by **AlphA** » 27 Oct 2005, 10:24

I also have an EM64T CPU , but I run on 32-Bit Win2k. Maybe thats the cause of my low scores in pi?

I dont know if you would get a better score on WinXp 32-bit.

jPm · Post by **jPm** » 27 Oct 2005, 12:22

XP 64 IS SUPER FAST COMPARED TO THE 32BIT VERSION,MAYBE YOU SHOULD TRY IT

Post by **Anakha56** » 27 Oct 2005, 12:28

jPm your caps is stuck again

excuse my ignorance but doesnt super pi first have to be written to take advantage of the extra headroom that 64bit offers? i personally feel it wouldnt make a difference as the proggi is written in 32 bit code so in XP64 it will work in 32bit. some1 please correct me if i am wrong...

AlphA · Post by **AlphA** » 27 Oct 2005, 18:27

Yes, I agree with you Anakha. The software would have to be written to take advantage of the 64-bit architecture (like the OS is) and compliled using a 64-bit compiler.

However, there might already be a 64-bit version of super pi out there....

BTW : Just for the hell of it, I ran Super Pi on my desktop at work (P4 2.4Ghz) and is scored a lousy 1m17ecs! I will run it on my laptop (Pentium M 1.7Ghz) a bit later, too lazy to go get it out my car in the garage!