[UPDATED] P5K WS + QX9650es Performance Test !!!!!!!

K.J. · 12-23-2007, 05:51 AM

I've a AVI of 512FSB boot of this super chips ...

Update 26 Dec. 2007

BIOS 0902 BETA too suxxxxx ..... I can do noting ... and I've many problems for rechange to 0805 ....

Update 25 Dec. 2007
Oh la la ~~~
90% of the time in tonight was passed on Air...2'C Ambiant ... o_O
And I've tested a BOX(I've just opened tonight)
I can't beleive that it has too strong FSB !!!

Max bench FSB 517.9MHz
http://francekj.googlepages.com/mygod.PNG

This speed for sub 10sec , is it fast ? lol
http://francekj.googlepages.com/best513FSB1M.PNG

And the 32M , limited by RAM

I can't vMOD the mobo ... so

I can't Sub 10min with 4.6GHz
http://francekj.googlepages.com/10m02.312s.PNG

The max 1M on air ....
http://francekj.googlepages.com/9.484s.PNG

And the max FSB on air
http://valid.x86-secret.com/show_oc.php?id=286717
@.@

BATCH just L739A643.....4A0104

Update 24 Dec. 2007

Now the only event I knew, that's C740 need(like)much of vCore ! in SS 1.52v 4.86GHz to enter the Windows but blue screen ! 1.6v I boot directly @ 5.06GHz no problem !! and 5.14GHz~5.15GHz Pi 1M
The max vCore of P5K WS will be changed when we change the cooling o_O on air 1.6125v BIOS = 1.632v Real.
In SS Cooling 1.6125v BIOS = 1.600v 1.6875v = 1.616v @.@
So I can't push to max for C740. and i rechanged to C738 .
C740 : Very big diffirence between 1.52v and 1.6v ... I told myself... It's possible that C740 is better than C738, isn't it ?
Finally I've done all tests today by the C738. because 1.6v is not fit for C740.it has the same fuction as C738 @ 1.52v
C738 @ 1.52v 5.04~5.05GHz 32M 5.14~5.15GHz 1M
@ 1.6v = 5229MHz 1M 8.843s just done tonight. Compared with 5145MHz @ 1.52V . 0.08v of changement = 84MHz progressed.
and 5120MHz 32M @ 1.6v 9m15.968s compared with 1.52v 5045MHz 9m21s -> 75MHz progressed

But C740 @ 4.86GHz 1.52v can't run into Windows, but 1.6v 5.06GHz run into Windows no problem.
200MHz of the diffirence ....

OK I post these screens of tonight.The mobo always no vMod. But I don't tweak the MemSet because it can limit very much frequence for my kit TeamGroup.
5.12GHz 5:6 512x10 32M 9m15.968sec almost killed the 9m13sec result when I was done in the LanParty of OCM Team....2seconds of the diffirence

http://francekj.googlepages.com/8.843s.PNG
http://francekj.googlepages.com/9m15.968s.PNG

And I've tested a Q9300es in SS. P5K WS BIOS 0805 don't support multiplier 1/2 ... so I can only use 7 max.
6MB La Cache for 4 Core, 1.5MB per core... compare with Q6600 Q6700 QX6850 8MB L2 Cache for 4Cores 2MB per core. Which is faster ?

http://francekj.googlepages.com/12m22.453s.PNG
http://francekj.googlepages.com/13.531s.PNG

Tomorrow I will retest the C740 on the GIGABYTE P35-DS4 who can give 1.6+ 1.7+ of vCore. for check the maxi CPU-Z in SS , for the 32M test, no .. not on the GIGABYTE.. I'm a idiot when I use GIGABYTE

But perhaps I will test 1 or 2 times of 1M....

============================23 Dec. 2007

P5K WS + QX9650es Performance Test !

I've tested 6pcs QX9650es, the 1st pcs can 1.280v 400x10 4GHz/1.472v 400x11 4.4GHz Dual Orthos Stable...
It's good .... lol

But this piece does not like too much of vCore,on the P5K WS mobo, 1.5875v=1.536v 1.6125v=1.632v direct..
and when it is working with 1.632v on Idle the processor is already 85'C o_O , I can't use 1.544v or 1.552v because of the BIOS ... Very Strange
So, all of my tests passed in 1.528v (Full Load 1.536v, that is the Legend of a great mobo...No vDroop at all !! )

MOBO: ASUS P5K WS Rev 1.02G P35+ICH9R BIOS 0805
CPU: Intel Core 2 Extreme QX9650es 333x9 3GHz C738C551 -43'C Load 32M
Cooler: SS by skirms_fr
RAM: TeamGroup DDR2-800 4-4-4-10 D9GMH 1GBx2
HDD: Seagate 7200.10 IDE 320GB 7200rpm 16MB Cache
VGA: Canopus G32 GeForce2 MX400 32MB DDR PCI
PSU: PC Power & Cooling TurboCool 1KW-SR

MACHINE
http://francekj.googlepages.com/machine.JPG

SuperPi 1M on air, Sub 10sec directly...easy !
http://francekj.googlepages.com/9.766sNoTweak.PNG

OK, Single Stage test . SuperPi 1M Sub 9
http://francekj.googlepages.com/8.984s5145MHz.PNG

When I test the 32M, I've met some problems ... Very Very slow, and I've ran 2 times 5.07GHz but always 9m50s... my god , Clear CMOS, and re-test...SS is already feeble

So , i've done a test of comparation of 5026MHz

it was funny, too .

386x13 2:3 RAM runing on 580MHz 4-4-4-4 MemSet Performance Level 4 9m38.125s
http://francekj.googlepages.com/5026...39m38.125s.PNG

9m38sec it's fast ? the result of 5:6 502.6x10 was increadible...Normaly Performance Level = 6 will be bery slow , but finally it was 15sec faster than 2:3 Level 4 !!!
http://francekj.googlepages.com/5026...09m23.765s.PNG

Then i've tested the max performance 5:6 , I would to sub 9m20sec. but the problem is the vDIMM of P5K WS, 2.55v = 2.522v Idle 2.517v Load 32M , my TeamGroup ran with 2.517v and the fastest timing, so they couldn't display the best performance(frequnce limited)
http://francekj.googlepages.com/9m21.953s5043MHz.PNG

4.8GHz compare the performance with 400x12 and 480x10

400x12 Performance Level 4 2:3 600MHz
http://francekj.googlepages.com/9m55...8GHz400x12.PNG

and 480x10 Performance Level 5 4:5 600MHz ... 11sec progressed !! It's the time to say goodbye for 2:3...
http://francekj.googlepages.com/9m44...8GHz480x10.PNG

Finally , i've tested 500x9 and 500x10 5:6 Level 6
4.5GHz 10m15sec it's not bad !
http://francekj.googlepages.com/10m15.453s4.5GHz.PNG

5GHz 500x10 Sub 9m30sec easily
http://francekj.googlepages.com/9m27.765s5GHz500x10.PNG

But 4:5 is faster lol 4907MHz 9m32sec JUST 5sec of diffirence between 500x10 5:6 and 490x10 4:5 !!!!!
http://francekj.googlepages.com/9m32.000s.PNG

Max FSB in SS Cooling 522MHz . 524MHz Freeze ...
http://valid.x86-secret.com/show_oc.php?id=285794

OK , all tests were done.

and i will give the solution of vCore setting to you.

in BIOS -> Actual voltage

1.6125v -> 1.632v
1.6000v -> 1.336v
1.5875v -> 1.536v
1.5625v -> 1.528v~1.536v
1.5500v -> 1.320v
1.5375v -> 1.528v
1.5125v -> 1.520v
1.5000v -> 1.304v
1.4875v -> 1.512v
1.4625v -> 1.504v
1.4375v -> 1.496v
1.3875v -> 1.488v
1.3625v -> 1.480v
1.3375v -> 1.472v
1.3125v -> 1.464v
1.2875v -> 1.456v
1.2625v -> 1.448v
1.2375v -> 1.440v
1.1000v -> 1.200v

Very Strange ... So we waiting for the new BIOS ...

OK , Thank you !

Pedro Rocha · 12-23-2007, 06:28 AM

KJ

The VCore of this mobo is really strange : 1.6000v -> 1.336v

did you measured with multimeter ??

Regarding performce, you have to count with strap changes - sugestion put 266 in FSB Strap in bios

Also remenber that with hight temps CPU may trotle down ..

Get a Foxcomm Mars and Shaminio will correct all bios bugs for you

K.J. · 12-23-2007, 06:32 AM

Quote:

Originally Posted by Pedro Rocha

KJ

The VCore of this mobo is really strange : 1.6000v -> 1.336v

did you measured with multimeter ??

Regarding performce, you have to count with strap changes - sugestion put 266 in FSB Strap in bios

Also remenber that with hight temps CPU may trotle down ..

Get a Foxcomm Mars and Shaminio will correct all bios bugs for you

It's real. 1.6v BIOS = 1.336v CPU-Z , have not tested by multimeter , but i can't do 4.3G 32M

the BIOS has not an option of Strap in BIOS

Oliver · 12-23-2007, 07:01 AM

Looks like you have been busy

.

What is the stepping of your 6 qx?.

Keep pushing it.

K.J. · 12-23-2007, 07:46 AM

3pcs C738C551 3pcs C740A520
I will teste a C740 tonight !

~~proxis~~ · 12-23-2007, 09:46 AM

zomg!

i like your subs!

CCUABIDExORxDIE · 12-23-2007, 09:49 AM

cpu-z volts usually mean nothing....id say bust out the multimeter and get some real volts.

pepinorang · 12-23-2007, 09:54 AM

Quote:

Originally Posted by CCUABIDExORxDIE

cpu-z volts usually mean nothing....id say bust out the multimeter and get some real volts.

Here, since CPU-Z 1.41 there's no big difference between multimeters and CPU-Z

Bye
p€p|n0r@ng

CCUABIDExORxDIE · 12-23-2007, 09:58 AM

how so..motherboard sensors can be off. i guess im just used to not being able to trust software for temp readings or for voltages.

redratamd · 12-23-2007, 10:09 AM

Very nice chips.
Please give one to me

Nice time too .

K.J. · 12-23-2007, 10:18 AM

Quote:

Originally Posted by Pedro Rocha

Get a Foxcomm Mars and Shaminio will correct all bios bugs for you

I just want a gift from him

K.J. · 12-23-2007, 10:19 AM

Quote:

Originally Posted by CCUABIDExORxDIE

how so..motherboard sensors can be off. i guess im just used to not being able to trust software for temp readings or for voltages.

When I was noting these voltages in the paper , I just read the voltage in BIOS , not CPU-Z

K.J. · 12-23-2007, 05:25 PM

updated

artkobe · 12-23-2007, 11:44 PM

guy!!!i want to rob your QX9650es!
and why not use DDR3?

K.J. · 12-24-2007, 04:15 AM

Quote:

Originally Posted by artkobe

guy!!!i want to rob your QX9650es!
and why not use DDR3?

My B.E. and P5E3 WS Pro is in Germany...

DDC · 12-24-2007, 04:35 AM

Nice!!!

before · 12-24-2007, 04:52 AM

Very nice stuff Kenny

Oliver · 12-24-2007, 05:11 AM

Cool update, looking forward to more.

Keep pushing it.

artkobe · 12-24-2007, 07:34 AM

Quote:

Originally Posted by K.J.

My B.E. and P5E3 WS Pro is in Germany...

XBX's?
let me send one to you ,hahah~
i just have B.E. and M.E.
i like P5E3 WS Pro,but asus didnt give me

K.J. · 12-24-2007, 08:13 AM

Quote:

Originally Posted by artkobe

XBX's?
let me send one to you ,hahah~
i just have B.E. and M.E.
i like P5E3 WS Pro,but asus didnt give me

No , they are not from him.

What will you send to me ?

artkobe · 12-24-2007, 07:37 PM

Quote:

Originally Posted by K.J.

No , they are not from him.

What will you send to me ?

Give me money

Miravo · 12-24-2007, 08:18 PM

Nice OC

And BEAUTIFUL memory, I am jealous

K.J. · 12-24-2007, 09:04 PM

Quote:

Originally Posted by Miravo

Nice OC

And BEAUTIFUL memory, I am jealous

I need to vMod ..pfffffffffff

pepinorang · 12-24-2007, 11:28 PM

SO it seems you finally find a CPU for me not as good as expected but pretty good

Thx...

Bye
p€p|n0r@ng

artkobe · 12-25-2007, 12:21 AM

great chips
and test the coldbug quickly

maybe 6.2G chip;D