Super pi 1M @3.6ghz

[shimmishim]

13.906

Abit IP35 Pro
E6600 @ 8x450
2x1 crucial tracers pc8000

[Fr3ak]

Too bad I noticed this thread a day too late. I just replaced my P5W with some crappy DDR3 board :/

I could run 4 3 3 4 with tweaked subtimings with the P5W. Can't remember what time I got, but it was 13.8xx something.

[BigJoe]

I was testing tweaks on my Win XP and got some interesting results
It is still not like my friend staxm result but...

I was burned my abit QuadGT

setup:
Core2 Duo E6850 ES@3600MHz, TT BT cooling
ASUS P5B-Plus Vista Edition
Kingston HyperX PC8500 2x512MB (D9GKX)
...

1M time=13.968s

[Chicco85]

This is my try...

Config:

- DFI 965-S (400*9, 2:3)
- Intel E6600
- Crucial ballistix tracer pc667 c3
- xp sp2 lite with some tweaks



Unfortunately, i made this shot above (before the other), but i could't open cpuz... it's too hot now in Italy, and the ram doesn't like it... I will try for a better results

[Bigchrome]

I finally got back from real life! The Scores Have been Updated guys!! Some of these are insane!!! I also have a new time of 13.891s:
13.891s (my windows install is getting old now! Maybe I'll borrow the drive from the HP downstairs that I just redid, 80MBs memory in use at idle!)

[stepsecr]

do i entitle to join the fun?

[Bigchrome]

No, it's called @3.6ghz for a reason.

[JMKS]

Strike again with Allendale .
It's not my 1M Windows (actually it's 16/32M etc) so I think it can be atleast 50ms better, if I will find time do to this (install another P35 Windows) I will post again .

Asus P5K-Dlx
E6400 450x8 = 3600MHz @ 1,45V
2:3 675MHz 5-5-4-4-4 @ 2,55V {end of V @ BIOS } Crucial Ballistix 8000 - P35 is making miracles with that RAM

15.219s




edit:
The Windows is still that non-1M-optimized 2003 (1M on XP is much better in my case) but the chipset is i965 this time - I love that chipset to be honest .

4:5 with modded 1333 strap (and from what I see that "modded 1333" is with 4:5 faster than genuine 1066 4:5 ) and a little better - P35 is slower with 4:5 about 1% than i965 4:5 so P35 3:5 5-5-4-4-4 seems slower than i965 4:5 4-4-4-4 .

[Chicco85]

Very small improvement....

Config:

- DFI 965-S (400*9, 2:3)
- Intel E6600
- Crucial ballistix tracer pc667 c3
- xp sp2 lite with some tweaks

[ownage]

Strike again with Allendale .
It's not my 1M Windows (actually it's 16/32M etc) so I think it can be atleast 50ms better, if I will find time do to this (install another P35 Windows) I will post again .

Asus P5K-Dlx
E6400 450x8 = 3600MHz @ 1,45V
2:3 675MHz 5-5-4-4-4 @ 2,55V {end of V @ BIOS } Crucial Ballistix 8000 - P35 is making miracles with that RAM

15.219s




edit:
The Windows is still that non-1M-optimized 2003 (1M on XP is much better in my case) but the chipset is i965 this time - I love that chipset to be honest .

4:5 with modded 1333 strap (and from what I see that "modded 1333" is with 4:5 faster than genuine 1066 4:5 ) and a little better - P35 is slower with 4:5 about 1% than i965 4:5 so P35 3:5 5-5-4-4-4 seems slower than i965 4:5 4-4-4-4 .

Nice OMG

When i have the possibility to reinstall Windows, i try to come below 15,4seconds. I dont have D9 ram, so i can only bench @ low ramspeed.

[Chicco85]

Now it's better:

[JMKS]

Nice OMG

When i have the possibility to reinstall Windows, i try to come below 15,4seconds. I dont have D9 ram, so i can only bench @ low ramspeed.

Thx , i965 with 4:5 really rocks {or 2:3 with P35, but as You can see i965 is still equal or even better }.

Sub15,4 450x8 1:1 could be difficult - I tried to re-achieve my previous 14,437 @ 450 1:1 today and I wasn't able to (another installation of Windows, maybe I changed something important ). Anyway, good luck!

Finally, I can say that this score "doesn't suck" as mentioned in first post of that topic .

Yes, that Ballistix I have for some tests is really good - 700 Dual Channel benchable @ 2,55V for example {on P35}.

[before]

Another one but 4.4.4
13.828s 3608MHz 2:3 4.4.4...
http://www.overclocking-jp.eu/public.../13.828s_2.png

Congrats my friend

[dinos22]

Another one but 4.4.4
13.828s 3608MHz 2:3 4.4.4...
http://www.overclocking-jp.eu/public.../13.828s_2.png

WTF
how the heck did you pull that time off

[Gautam]

3608

It makes a large difference...I did this originally but never wanted to post it because I believe that 3.6GHz=3600MHz and no higher. But that's just me.

koda's time is the most efficient so far in this thread, barring none, and that no one can argue with.

[v_parrello]

Here there is a system DDR3 powered.
Asus P5k3
E6420 air cooling
DDR3 SuperTalent 1600 7-7-7 Micron Z9
Win2K3

[Learn]

Congrats my friend

thx

WTF
how the heck did you pull that time off

Some basic tweak, wich everybody know now

3608

It makes a large difference...I did this originally but never wanted to post it because I believe that 3.6GHz=3600MHz and no higher. But that's just me.

koda's time is the most efficient so far in this thread, barring none, and that no one can argue with.

http://www-personal.umich.edu/~gautamb/3607_13828.PNG/

See posts #1 #72 #74 #92
no
There is someone (i know) with a much better efficient and a italian dude also

[Gautam]

Hm excuse me Learn I haven't been keeping up with this thread. But koda still is .001 faster than you.

I know the Italian time you're talking about....I completely do not understand it.

[Learn]

Hm excuse me Learn I haven't been keeping up with this thread. But koda still is .001 faster than you.

I know the Italian time you're talking about....I completely do not understand it.

I would like to know too there is a long road it seems

[shimmishim]

oh sweet oogly moogly.. i'd like to see this uber sweet italian's score!

but Learn, nice clocks and speed!

it's very good to see early samples of DDR3 doing so well.. then again I guess it was like early DDR2 with Micron fatbodies...

[Gautam]

oh sweet oogly moogly.. i'd like to see this uber sweet italian's score!

By C.h.r.i.s-



http://www.xtremesystems.org/forums/...0&postcount=33

Completely incomprehensible. Even 13.828 is very difficult at 3603. This 781 is a good 3 ticks ahead of any other time in this thread.

[Stapler]

That time really doesn't make sense, I gotta wonder if he's hiding the memory info for a reason, and also what the available system cache was during that run (what makes copy waza faster). The RAM timings are tight, but not tighter than any one elses, and not enough to make a difference, so that can't be it.

[Gautam]

The available real memory is likely part of the key. (See his first 1M result in that post where he shows it)

[StealthyFish]

rithina its easy on commando, to go over 600 , just give your ram 2.95v and thats the score you get
http://valid.x86-secret.com/show_oc?id=189806

haha. he makes it sound so easy. Only 2.95V

[Timbosan]

That time really doesn't make sense, I gotta wonder if he's hiding the memory info for a reason, and also what the available system cache was during that run (what makes copy waza faster). The RAM timings are tight, but not tighter than any one elses, and not enough to make a difference, so that can't be it.

Did you actually look at anyone else's timings lol? The ram timings are actually tighter than most here - he's at 4-4-3-4, the TRFC is 21 as opposed to 30 - 45+ and write to precharge & tWTR are both 9, and TRRD is 1.