AMD X4 940 OC test in MSI DKA790GX-Platinum !!

[Toppc]

DIMM x4 TEST 2
2GBx2 + 2GBx2

Transcend PSC AA3G 2Gx2
+
Corsair CM2X2048-8500C5D
TOTOL 8GB

DDR2-1120 CL5-5-5-15 TRFC 195ns

[accord99]

wow that looks really good. Thank you for taking the time to post results. Much cooler then Intels Current Quads. I just might move back to AMD. Support the little guy.

The power measurements indicate the AMD temperature and Intel temperature readings aren't comparable.

[Toppc]

BUS SPEED 350
OC 4.2G CPU-Z DOWN

[Oliverda]

use XIGMATEK HDT 1284C
NO OC , VCC 1.35V
Temperature Descend

1 SP2004
almost 40W


2 SP2004
almost 55W


3 SP2004
almost 70


4 SP2004
almost 82


WAIT 10 MIN
4 SP2004
almost 88W

Why did you get lower values with bigger cooler? If I'm not mistaken the power consumption doesn't depend on the cooler.

What kind of PSU did you use for the test?

[Caveman787]

I'm gonna guess less heat less power consumption needed.

[Zucker2k]

I'm gonna guess less heat less power consumption needed.

Yeap.

[Oliverda]

I'm gonna guess less heat less power consumption needed.

Less power consumption ---> less heat. It is not true vica versa.

[Opteron146]

Less power consumption ---> less heat. It is not true vica versa.

It is true, the better you cool the less power the chip needs.

cheers

Opteron146

[Oliverda]

It is true, the better you cool the less power the chip needs.

cheers

Opteron146

Sorry but it's nonsense. If it would be true we can get approximately 1W with LN2.

[Caveman787]

No...it still requires power but less power is wasted by the taxing of power control such as in pwm and mosfets.

Atleast that's my limited understanding take.

10 posts since 2004 wow that's control.

[ryboto]

It is true, the better you cool the less power the chip needs.

cheers

Opteron146

Increase the heat, and you decrease efficiency. So, it makes sense cooling the chip better should slightly decrease power consumption.

[MadDias]

i´m on oliverda´s side on this one...
heat is generated by the processor. with a bigger (better) cooler it´s still drawing the same amount of power. efficiency has nothing to do with the heat.

[crazydiamond]

maybe why the guys on air are taking more v's to get to 4ghz than guys on water/cold?

[MadDias]

maybe why the guys on air are taking more v's to get to 4ghz than guys on water/cold?

that´s another story. stability isn´t equal to efficiency.

[Sgt.McRuff]

I thought that the colder the chip the less transistor gate leakage there is. Chips still need same amount of power, but less is lost in achieving it when cooler.

[Manicdan]

can you run cinebench or 3dmark at 208x18 and 340x11 (both will be around 3740-3744 mhz), with the RAM at about the same speeds and timings. i want to see what kind of bonus we can see by maximizing the bus speed with almost everything else the same

[justapost]

I thought that the colder the chip the less transistor gate leakage there is. Chips still need same amount of power, but less is lost in achieving it when cooler.

It's like you said.
@Oliverda, you can test it with a power meter and an fanmate. You can easy squeezup power consumption by reducing the fanspeed, that's why I do not care much about power consumption reviews without temps mentioned.

+1 @ toppc, good testings there. Can you run 3GHz at 1.185V like hondaguy did in the other thread?

[Toppc]

Why did you get lower values with bigger cooler? If I'm not mistaken the power consumption doesn't depend on the cooler.

What kind of PSU did you use for the test?

ENHANCE 650W

Test process in.Only change fan

[Toppc]

It's like you said.
@Oliverda, you can test it with a power meter and an fanmate. You can easy squeezup power consumption by reducing the fanspeed, that's why I do not care much about power consumption reviews without temps mentioned.

+1 @ toppc, good testings there. Can you run 3GHz at 1.185V like hondaguy did in the other thread?

This place is evening
I need sleep
Wait for tomorrow ok ?

[G0ldBr1ck]

Sorry but it's nonsense. If it would be true we can get approximately 1W with LN2.

As the temp increases circuit resistance increases, more energy required to power the circuit.

[justapost]

This place is evening
I need sleep
Wait for tomorrow ok ?

Of course.

[TL1000S]

Thanks for sharing Toppc.
Mucho Vcore, so the temps were not that bad I think.
Great CPU-score in 3DM06 also.
If this is representative for what we will get of retail PHII, this will be fun!

[Lightman]

Very good work Toppc!

Do any of you have data regarding PWM efficiency on motherboard (especially this MSI)? We would be able to work out actual power consumption of CPU that way.
Anyhow it looks like using this stepping AMD if wanted could release one more speed bin @3.2GHz and still be under 125W and with 140W 3.4GHz should be also possible.

Besides you need to love this CPU with C'n'Q enabled idling @1GHz/1V
Can you measure power consumption at above mentioned setting Toppc?? Please

[gallag]

Thanks for sharing Toppc.
Mucho Vcore, so the temps were not that bad I think.
Great CPU-score in 3DM06 also.
If this is representative for what we will get of retail PHII, this will be fun!

Obviously the temps are not being reported correctly. Still good clocks though and if it is stable for a few hours the temps are not the most important factor. I defo think that denebs temp sensors are not as accurate as i7's.

[G0ldBr1ck]

Obviously the temps are not being reported correctly. Still good clocks though and if it is stable for a few hours the temps are not the most important factor. I defo think that denebs temp sensors are not as accurate as i7's.

or i7 sensor is not as accurate as Denebs. Its all speculation so can go both ways right? what is the basis for determining which is more accurate?