DI-party in Warsaw, PCLab.pl involved :), E4400 & Cel420 WRs :D

[JMKS]

Hello!

Last weekend we had a DI-party* in Warsaw: , including a "session" with E6750 ES in the PCLab.pl headquarters .

*my first-ever DI and second subzero - the first was with piotres's cascade - http://www.xtremesystems.org/forums/...d.php?t=142898, the results {not great for an E6700 L629B, time and VCore limited} http://www.xtremesystems.org/forums/showpost.php?p=2172518&postcount=54}

Hardware:

Asus Commando + VDroop [pencil] {"both"} by JMKS
VGA 8600GTS by PCLab.pl, 8500GT DDR3 by jumson, Gf6600 by JMKS {no graphic benches }
PSU Silverstone 850W by PCLab.pl + ModeCom MC500 by JMKS
RAM Corsair Dominator 10000 by PCLab.pl + GoodRAM PRO by JMKS
CPU's: E4400 by jumson, E6750 by PCLab.pl {or by Intel, I don't know what to write }, E6400 by JMKS & Cel420, I don't now who is the owner of this (me or my friend takero, we must decide afterwards )
Thermal grease Arctic Ceramique (thx Team Japan for the "tip" )
DI container by piotres


Scans of the CPU's used:

Cel420


E4400


E6400


E6400 "1333 FSB mod" - unfortunately it isn't working at all (CPU is working but 266 FSB standard, nothing else) - maybe the "graphite ink" used was not that special, hmm...

Container:


Some pictures of the setup (more to come!):






Some photos from my cellphone (as I said before, more high-quality photos will be there shortly ):




[img]http://img507.imageshack.us/img507/677/dsc00584ob7.th.jpg[ /img]
cold


after the tests with E4400, WR done! {easily }

and the best:

shortly after the benches, the view from my windows - as You can see the sun is shining - I started this "session" shortly before midnight and tha photo was taken at 4:36 AM, I'm sleepy...
czyli widok z okna tuż po zakończeniu benchów - jeśli wierzyć komórce, to zdjęcie zrobione o 4:36 AM - coś wcześnie / późno (zależy jak patrzeć ), w każdym razie zważywszy na to, że 2 dni wcześniej miałem 36-godzinną dobę to będę to musiał niedługo odespać :]


After the session with Cel420 {another WR }, some freezing action {the PC isn't running of course}




After all the tests. In this picture You can see the homemade heatsinks for MOSFET cooling


OK, let's talk about the results :

First test was with E6750. It is great CPU for daily use, 4GHz fully stable possible with ~1,4V I believe . Air pretesting gave me some good results - 4080MHz 1M with 1,55V (lower not used ) and SetFSB to 515 FSB => freez.

Under DI the FSB-wall was absolutely frustrating - highest benchable FSB was 520 which gives 4160MHz "only". BTW, "strap modification" was used {thanks to jp @ ixtremtek.com I must say }- no genuine 520 with air cooled NB and my Commando possible.


Max validation: http://valid.x86-secret.com/show_oc?id=213054


4GHz {pseudo, only 15mins but who is doing Orthos under DI?!}-stable with 1,35V - I don't know if it is great or "only" very good .


Now E4400, really sweeeeet :

Max air, 1,70V, 4090MHz

http://valid.x86-secret.com/show_oc?id=212166

Picture of the air setup (it's my 24/7 setup , of course except of CPU, and VGA):





Max 1M 4010MHz 13,89s - if someone's interested: http://hwbot.org/signature.img?iid=25007&thumb=false

Now let's cool it "a little" :

all results with 1,7625V @ BIOS => 1,808V idle in BIOS, slightly higher under load I think - the result of the VDroop
FSB Termination 1,45V
NB VCore 1,75V or so - with HR-05 this NB survived 1,95V for a lot more than "couple" of hours , 1066 strap in all the results of course

max MHz - 4710MHz - WR by more than 200MHz I believe
FSB-wall limited maybe, 475x6 => no POST, FSB-wall on air ~440

http://valid.x86-secret.com/show_oc?id=213661

max 1M - 4620MHz - 12,375s - not bad for an Allendale IMHO


16M "Polish national bench" {but not maxed, You will see }, 4,53GHz


32M with higher frequency than 16M , 4,54GHz


The clock on the screen is working correct, yep, it was close to 4AM

And a big disappointment - PiFast wasn't stable even with 4,4GHz - WTF?! 4,35GHz finally:


The end of the E4400 benching and I could not resist, once more:
short after the bench ends


The next day and the new CPUs to test :
Celeron 420 with hmm... 1,45V {overkill! , stable with 3,2GHz and 1,35V on air}
On air the FSB wall even for validation is ~425.

max MHz: 3552MHz - WR (temporary )

http://valid.x86-secret.com/show_oc?id=213664


In the PCLab.pl headquarters I also had chance to bench WPrime on the dual Clovertown machine - 7,66s 32m and 3m44,98s in 1024m.

More photos to come! I hope You like this .

Enjoy!

{PS. It seems that I must polish this in some places because the photos etc isn't showing correctly - give me a moment and it will be OK - that's the result of using notepad for writing }

[G.Foyle]

Nice results, congrats for WR!
Sam bym chętnie pobenchował DI ale brak mi tuby

[Viss]

Very nice E4400! That E6750 needs some work hehe. What vcore you used on that proc?

[JMKS]

Nice results, congrats for WR!
Sam bym chętnie pobenchował DI ale brak mi tuby

Dziękuję / thx . Sprzęt do DI jakiś kosmicznie drogi nie jest .

That E6750 needs some work hehe. What vcore you used on that proc?

All range: 1,30, 1,35, 1,4, 1,55, 1,6, 1,7, 1,8. FSB Termination 1,20, 1,30, 1,35 (usually), 1,45, 1,55. PCI-E 90, 120. NB VCore 1,25, 1,55, 1,75. SB V 1,50, 1,60, 1,70 - nothing.
FSB-wall and "You can do nothing" (maybe LN2 but with that kind of wall it is in fact useless) .
But E6850 seems not to have that low walls - maybe soon I will have chance to test one .


Some "xtreme" was when on DI I decided to downgrade BIOS from 1001 to well-known and most-used 0803 . No problems fortunately but no improvement...

[DeDaL]

About 6750, FSBWall here ?

E6750 seems is not for 5Ghz+

[JMKS]

Yes, exactly - FSB wall...
Team Turkiye catched waaaaay better E6750 - http://valid.x86-secret.com/show_oc?id=210128 but judging by some reviews on the Internet of the E6750 the ~4GHz (probably walled too) is the limit of nearly all E6750. Maybe retail will be better.

And this CPU has FSB-wall on air with P5K-Dlx on ~490 (tested not by me) - I don't really know why on Commando (and air, too) it was "simple" 510+.

[Pt1t]

Yes, exactly - FSB wall...
Team Turkiye catched waaaaay better E6750 - http://valid.x86-secret.com/show_oc?id=210128 but judging by some reviews on the Internet of the E6750 the ~4GHz (probably walled too) is the limit of nearly all E6750. Maybe retail will be better.

And this CPU has FSB-wall on air with P5K-Dlx on ~490 (tested not by me) - I don't really know why on Commando (and air, too) it was "simple" 510+.

MY E6750 is also stable @ 4ghz , But on aircooling

http://www.pt1t.net/E6750/Prime_500X8_Air.png

And MAx FSB is 530 on Air Also.

[JMKS]

Tested 3,8GHz on stock V on air - stable, further was FSB wall (P5K-Dlx used). I think that 4GHz Orthos on air also possible and not hard to achieve - this screen was to "compare" that CPU to others - not many can do 4GHz Orthos @ 1,35V on DI I think .

I have not tested this CPU much on air - "2 freezes" and it goes for DI .

What VCore on Your 4GHz?

[DEVIL K-ce]

JMKS You know Good job

Congrats and go for more in next di party.


Martin

[Oliver]

Congrats with the results

Keep pushing it

[mac11]

Good ,good work , I wait on more more far , higher results!!!.

Last weekend we had a DI-party* in Warsaw: , including a "session" with E6750 ES in the PCLab.pl headquarters

maybe more dry ice and acetone haha and new delta - then joke
hm.

Yes, exactly - FSB wall...

bad.bad luck

Ciekawsze fotki niż na labie hehe (z wazonem ,drobniakami i kartą w tle) - tak a propo testów z Kontener-em
(oczywiście w pozytywnym ,słowa ciekawsze znaczeniu - pogląd na konstrukcję )
p.

[bedlamite]

Congratz guys. Go for more.

[FatAlbert]

nice score

PS.
kiedy powtorka?

[JMKS]

nice score

Thx .

kiedy powtorka?

When it's done .
With E6750 probably never (it isn't worth it ), with E6400 (no good score achieved) and Cel420 too, with E4400 maybe in a month or so.
Some other CPUs: E6850 ES, E6700 L629B (the same as with cascade) or X6800 (B1 unfortunately but PCLab has 2 X6800's) will be tested shortly i think .

The article @ PCLab.pl (many good photos, even 2 videos ): http://pclab.pl/art27275.html

[PcCI2iminal]

Nice 4400 and 420 you got there

congrats for the WR

[Turrican]

nice results guys.

[AragornPL]

Congratulations!
E4400 i Celek naprawdę fajne wyniki, gratulacje rekordu
A ten E6750 mógł być taki słaby dlatego, że to ES - nie wszystkie ES porządnie się kręcą, może retailowe będą lepsze?
Niezbyt dobrze zabezpieczyliście mobo, ale to szczegół - ważne, że nic się nie stało
A i Delta była dobra

[ent]

Nice results guys, especially those E4400 Pi runs

[SoF]

that e4400 is sweet. nice tube you got
really hope I can pick a similar chip when I try it under phase with some low cost e4XXX (or e21XX maybe for 4 gig).

[DEVIL K-ce]

Hey JMKS and all

Today also I received E6750ES and the same bad cpu.

FSB wall at 500MHz 499MHz is ok, but set 500MHz FSB and systems crashed. Tried any mem, any timinigs, any voltage of nb/sb/fsb ... and cooled this cpu @ cascade Nothing.

http://valid.x86-secret.com/show_oc?id=215171


So, no bench at this low clocks.


Martin

[hausner]

Hey JMKS and all

Today also I received E6750ES and the same bad cpu.

FSB wall at 500MHz 499MHz is ok, but set 500MHz FSB and systems crashed. Tried any mem, any timinigs, any voltage of nb/sb/fsb ... and cooled this cpu @ cascade Nothing.

http://valid.x86-secret.com/show_oc?id=215171


So, no bench at this low clocks.


Martin

Bad luck Martin

[JMKS]

QX6850 arrived yesterday, so it was nice to test this one .
With 4 cores enabled it wasn't that special but with only 1 it was quite good . 8800Ultra Gigabyte used for some 3D fun .








I'm still at the PCLab.pl headquarters ATM so I will write some more shortly, probably tomorrow .

About the last one (80741 3DMark2001, stock cooling VGA) - @Devil, Polish 1st non-SLI 2001 result or not, I'm not sure?

Max MHz:

http://valid.x86-secret.com/show_oc?id=215218

Enjoy!

[DEVIL K-ce]

Nice one JMKS

Go for better cooling and ...

Yes You beat my 01 score - congrats

4 core and dry ice hmm no go with high clocking, but Your single core score is ok.

So .. maybe small 3D competition

Martin