PDA

View Full Version : SSE2 Patched super pi for a64



OV3RCLOCK3R
04-16-2005, 10:38 AM
where can i get this from?

craig588
04-16-2005, 03:38 PM
Here, use this version so people arn't mad. It has a edited titlebar so people can tell it's patched, but you still get the higher scores.

OV3RCLOCK3R
04-17-2005, 03:21 AM
Here, use this version so people arn't mad. It has a edited titlebar so people can tell it's patched, but you still get the higher scores.

thanks matey:)

[XC] 4X4N
04-22-2005, 01:03 PM
I just downloaded this and ran a comparison with the mod 1.4 version. I went from 35.359 seconds to 34 seconds with this version. Is the patch a cheat to get a better score? Which one is a more accurate benchmark?

craig588
04-22-2005, 01:10 PM
It's just different. You get the same result and your CPU makes the same calculations, they are just calculated differently. It is faster because it uses more effeciant instructions. People don't like it because it has changed a long time standard of a benchmark and the first patched versions didn't change the title so it basicly was a cheat because you could say you used the regular version when you really used the patched version.

This one has an edited title so you can tell what version someone used. (You could always edit the title with photoshop, but if you wanted to do that you could have just edited your time)

It also uses a seperate database so you can keep your patched and standard results seperate.

glock19owner
04-22-2005, 01:41 PM
Is there a SSE2 patched version that reads milliseconds?

craig588
04-22-2005, 01:57 PM
Nope. Theres no real reason for that, if you want to go for WRs use the real one.

[XC] 4X4N
04-22-2005, 02:35 PM
It's just different. You get the same result and your CPU makes the same calculations, they are just calculated differently. It is faster because it uses more effeciant instructions. People don't like it because it has changed a long time standard of a benchmark and the first patched versions didn't change the title so it basicly was a cheat because you could say you used the regular version when you really used the patched version.

This one has an edited title so you can tell what version someone used. (You could always edit the title with photoshop, but if you wanted to do that you could have just edited your time)

It also uses a seperate database so you can keep your patched and standard results seperate.

Ok, thanks. I will keep both, but I like getting 34secs, though :D

glock19owner
04-22-2005, 02:48 PM
Nope. Theres no real reason for that, if you want to go for WRs use the real one.

Actually yes there is actually...I dont care about WR's...but I do want to know how close I am with my tweaking...

bachus_anonym
04-22-2005, 02:53 PM
Actually yes there is actually...I dont care about WR's...but I do want to know how close I am with my tweaking...
Forget about it, bro :) SSE2 / SSE3 patched-versions should have never existed in my opinion! Just stick with 1.4mod, ok?
We do not need any more confusion around this nice benchmark...

craig588
04-22-2005, 05:18 PM
Thats the exact reason. If I add anything extra to this version it will make it easier for someone to fake it as if this was the legit version. This is as far as it's going. The only reason I edited the title and released this version was because I knew that if someone didn't make a easily avalible version that shows you it's patched people would go out of their way to find a patched version without any indications of fraud.

If I added the Msec support it would just make a patched version that much easier to fake as a real one.

You can still see the differences in performance with the regular Msec version, the only reason someone would need them in this version would be for faking results.

Gorod
05-19-2005, 11:29 PM
how about SSE3 pached version guys ? I want to see if their any difference in performance :) Thanks you :toast:

x714n
05-19-2005, 11:42 PM
how about SSE3 pached version guys ? I want to see if their any difference in performance :) Thanks you :toast:
http://xtremesystems.org/forums/showthread.php?t=62806

Gorod
05-20-2005, 08:43 AM
http://xtremesystems.org/forums/showthread.php?t=62806

Thanks for reply but i dont see where to download it :confused: Tried to find it by google , yahoo and others and cant :(

x714n
05-20-2005, 02:49 PM
Thanks for reply but i dont see where to download it :confused: Tried to find it by google , yahoo and others and cant :(

http://i4memory.com/AMD64_tools/superPi/super_pi_SSE3.zip
get it from i4memory.com

Gorod
05-21-2005, 08:27 PM
http://i4memory.com/AMD64_tools/superPi/super_pi_SSE3.zip
get it from i4memory.com Thank you so much !!! :toast: :woot:

ewitte
05-23-2005, 06:02 AM
This one has an edited title so you can tell what version someone used. (You could always edit the title with photoshop, but if you wanted to do that you could have just edited your time)

Its really easy to fix the title with a Hex Editor ;)

NickK
05-27-2005, 06:11 AM
Sounds perfectly ok for benchmarking.

Benchmarking is about completing a set task as fast as possible, in the most efficient & optimal manner for your machine.

So running a recompiled benchmark that does exactly the same task and comes out with the same answer but does it in a more efficient way is fine by my book. Can't see why people really want to stick with an old version that doesn't use the new features of a CPU that you originally bought it for...

If we did that, we'd still be using 8086 code without SSE or even 32bit integers..

Bring on 64bit, bring on SSE3, bring on multithreading and dual core and lets see who completes the task the fastest.

DudeMiester
05-27-2005, 02:56 PM
Well I think that's a good idea, but it would have to be a totally new bench. Maybe call it "Ultra-PI" lol.

craig588
05-27-2005, 04:29 PM
Its really easy to fix the title with a Hex Editor ;)

I guess that must be one of my early versions. One of the later ones I edited would crash if you changed the title. (Intentional crash, not just a bug)

ewitte
05-29-2005, 10:09 AM
I guess that must be one of my early versions. One of the later ones I edited would crash if you changed the title. (Intentional crash, not just a bug)

If you delete it crashes. Just put a space for each character your commenting out. It has to be a hex editor. Notepad/wordpad/edit will not cut it.

craig588
05-29-2005, 12:27 PM
No, I mean that you couldn't change the title at all. I rearanged how it looked for the title.