The Fastest In The Land.

I would LOVE to see a Mac Pi program <_

#45021

Originally posted by Taha@Aug 5 2003, 02:17 AM

I would LOVE to see a Mac Pi program <_<

Wait - Dont you mean APPLE PI ???

:lol::lol::lol:

I'm going strong tonight...

#45022

LOL. Apple PI. :lol:

DZ, you make a great drunk.

Taha:

http://www.myownlittleworld.com/pi/program.html

Works in OSX.

#45023

2.53 p4 533FSB 512 pc2700@333mhz

and DZ still has me, DOH, must be the ram

#45025

Woah, The AMDs are really pulling ahead. Im impressed. I guess this is why AMD is known for math calculation and Intel for 3D.

#45027

Before we all jump to conclusions, I just want to let you know there are various algorithms that can be used to find Pi.

There's the point-line-mode, Alternating Fractions mode, Brent-Salamin Algorithm and the Arc Tangent Algorithm.

Each algorithm yields different results. Now I don't know which one you're using, but using the app Contra mentioned I get these results (view attachment below). I suggest we all use this app since its cross-platform and perhaps we can have a better way to compare the various platforms as well our individual machines.

edit: Can anyone please try it out? I want to try it on my PC, but don't understand how this thing works!

#45030

I tried, but have no idea, what I am supposed to do with this.

#45032

Just do C:\....\desktop\Pi\blahblah.exe 1000000

Shît, my PC beat my mac by more than 10 seconds.. :(

#45034

Wow, a little bit of a command line and you windowers go and run. :lol:

In a Console, type pi_css5.i686.exe [amount]

I did 1000000

pi2.JPG

Anyway, this isnt that accurate, because there are no CPU specific optimizations, just raw CPU isnt good enough, we need SSE2, altivec, etc.

EDIT:\\ Im gonna try it in Linux later too. That way we'll see how accurate this thing is (theoretically, it should be the same).

#45035

On a 1.25GHz 1G RAM Dual G4, these results were achieved:

20.64 seconds  Dual 1.25 G4  

1 gig ram

I guess a G4 DOES make alot of difference for number crunching, huh?

on another similarly-equipped system, it took 14.8 seconds.

#45039

command line what is that???? :woot:

anyway here ya go

thanks contra and taha

sorry I shoudl probably mention this is a 2.53ghz 533fsb p4 512mb pc2700 ram @333

#45040

Ha! My P4 2.0Ghz scored as well as a dual 1.25Ghz G4 w/ a Gig of RAM :lol:

But thats not accurate. This little program hasnt optimized for Altivec, SMP, etc. But for single processors, it should be OK.

This seems more accurate than that other program though. Photos 2.4Ghz is (rightfully) scoring a lot better than my 2.0Ghz.

#45041

We have done some more updates to the Barton Overclocking Database and things are looking better now for Bartons. Apparently a new world record has been set with XP 2800+ (2.08Ghz) overclocked to 3.46Ghz and its [superPI 1M] calculation (32s) is completed faster than a P4 3.06Ghz@4.6Ghz (33s).

#45044

EDITED

#45045

Photo678: Great number crunching! My hat is off to the mighty P4!

Oh yes, the Athlon is no slouch

Pi_prompt.jpg

In case you forgot:

AthlonXP 2400+ (2.04Ghz)

I'm dying to see how well the new PPC 970's do, as well as the 3200+ Bartons with 512k cache, or even one of the newer 800mhzFSB P4's...

#45046

And since we are all insane, here's an icon for your app:

Pi_character.png

I know you guys are looking for something like this!

The icon was created by Pixeljerk, and I do have permission to distribute his icons.

#45048

hahaha, thats awesome

Thanks

#45049

Honestly, I feel we should be doing big numbers, like 10,000,000. It would be more accurate, and its not like it would take up too much of our time.

EDIT:\\ DZ: Great Icon. You deliver once again.

I just realized we're dorks.

#45055

ok ran it again at contras suggestion

#45060

Originally posted by Vejitarot@Aug 4 2003, 09:43 PM

Well I DO trust Norton. And I'm using Norton because it is easy to compare two machines with it.

Why does someone always have to start some crap? :rant:  j/k  But seriously, just take the post as it is and don't throw semantics into it. I offered Norton Performance Test files as a comparison basis so take it at that. Don't make everything more complicated than it has to be. That's the problem with this world, everybody makes everything so complicated when it's really simple. If you don't want to benchmark your machine using Norton...then don't. Just thought it would be fun to compare machines with it.

Don't guve a fukk about Him :):D Everything is kwel :woot: You re right!

#49786