Enormous heat diffrence between 2 equally specced mobos

386-sx-33 · Post by **386-sx-33** » Sun Jun 24, 2007 5:33 am

http://images.tweaktown.com/imagebank/n ... usttst.pdf

ASUS P5K Deluxe was high on my buying list until I found the above article.
The board gets up to 64C hotter than the GIGABYTE P35-DQ6
The very high temps won't do much good for component life.
Asus being a solid brand and the fact I can't find any other reference to this extreme temp difference makes me think the measurements are flawed.

For silent cooling it's important to know if that heat is just caused by bad heat sink and heat pipes; or that the Asus wastes a lot more power.
In the last case more heat need to be moved out of the case, meaning more noise.

Can anyone please shed some light on this article?

LuckyNV · Post by **LuckyNV** » Sun Jun 24, 2007 6:53 am

seen it before, personally i think it is propaganda until i see a proper test carried out by a reputable review site and with other motherboards.

Das_Saunamies · Post by **Das_Saunamies** » Sun Jun 24, 2007 10:40 am

That thing looks more like an amateurish advertisement than any real test paper.

Well, reading the whole story helps understand why: http://www.tweaktown.com/news/7496/index.html - story short: propaganda, maybe even lies?

Edit: Check these very forums, we have first-hand user experience: viewtopic.php?p=346429#346429. Quite ordinary temperatures, no doomsday inferno here -- quite the opposite.

So what have we learned about reading a competing manufacturer's "articles"...?

jojo4u · Post by **jojo4u** » Sun Jun 24, 2007 12:19 pm

I have evidence proving it. The Asus needs a massive amount of energy: viewtopic.php?p=349921#349921

386-sx-33 · Post by **386-sx-33** » Sun Jun 24, 2007 1:10 pm

C'T is one of the best printed PC magazines; so I put some trust in their numbers.
34Watt extra. I wonder why.
Just crappy design? Or because it has all kind of extra's ?
Can 34Watt generate such a heat diffrence?

386-sx-33 · Post by **386-sx-33** » Sun Jun 24, 2007 1:20 pm

Das_Saunamies wrote:
Well, reading the whole story helps understand why: http://www.tweaktown.com/news/7496/index.html - story short: propaganda, maybe even lies?

.....

So what have we learned about reading a competing manufacturer's "articles"...?

I guess the Gigabyte marketing drones where overheating.....
Perhaps they used SPCR measuring methods.
Asus measured at 30cm and gigabyte at 100cm

merlin · Post by **merlin** » Tue Jun 26, 2007 11:14 am

386-sx-33 wrote:C'T is one of the best printed PC magazines; so I put some trust in their numbers.
34Watt extra. I wonder why.
Just crappy design? Or because it has all kind of extra's ?
Can 34Watt generate such a heat diffrence?

34 watts more is a lot on a motherboard actually, especially since a lot of areas are not always well cooled. Well it's pretty noticeable that asus has a very different power design, I guess perhaps the 12 phase setup is causing more power loss than you'd expect...along with other unknown things.

386-sx-33 · Post by **386-sx-33** » Tue Jun 26, 2007 11:33 am

Asus has 8 phase power and Gigabytes 12 phase power.
Both claim it reduces power use.
Another thing is that the P5K3 (DDR2 and DDR3) uses 10 watts extra compared to P5K (DDR2).
Maybe that's why the P5K3 has an extra heatpipe.

Well the most important thing is that the extreme heat issue is just a lie.
But I agree that 34W is a lot. Especially percentage wise.

nick705 · Post by **nick705** » Tue Jun 26, 2007 12:20 pm

The latest group test at The Tech Report seems to confirm the Asus boards being much less efficient than the competition - consuming nearly 40% more power at idle than the equivalent Gigabyte...

merlin · Post by **merlin** » Tue Jun 26, 2007 6:38 pm

386-sx-33 wrote:Asus has 8 phase power and Gigabytes 12 phase power.
Both claim it reduces power use.
Another thing is that the P5K3 (DDR2 and DDR3) uses 10 watts extra compared to P5K (DDR2).
Maybe that's why the P5K3 has an extra heatpipe.

Well the most important thing is that the extreme heat issue is just a lie.
But I agree that 34W is a lot. Especially percentage wise.

I've never even owned a gigabyte board, but the extreme heat seems to correlate with the huge amount of power that asus boards are using. It may not be to the extent gigabyte is saying, but it's quite clear that temps on asus boards will be higher. I'm not sure how it's a lie per se, probably a very manipulated result that does not show normal conditions, but a real result nonetheless. Anyways nobody should listen to a report like that, just find a good trustworthy review site for REAL answers. (such as this one here!) The tech-report graph along with numerous others has given me one clear conclusion though, avoid asus p35's at all costs.

RBBOT · Post by **RBBOT** » Thu Jun 28, 2007 4:51 am

I think part of the problem may be the fact that the ASUS boards seem to pointlessly raise voltages left right and centre when in the default "Auto" configuration.

I've just got a P5K Deluxe with an E6600 in an Antec P182 with a scythe ninja and Noctua 800RPM fans on the two exhausts and on the scythe which I am currently letting run full speed. I'm hoping to take the fan of the cooler evenutally but day 1 i'm starting with active cooling. I have 4 x 2Gb DDR2-800 sticks of RAM in there and the moterboard temperature sensor is directly above it so I'm expecting fairly warm system temps.

Leaving all the BIOS settings on Auto, and running orthos blend mode with no overclocking I got temps: CPU 54, Core1 64, Core2 64 and system temp of 53. The MB had boosted vcore to 1.38. Even weirder when I stopped the test and left the machine idle, the speedstep sensibly brought the CPU multiplier down from 9 to 6, but the vcore increased to 1.395, probably due to no droop at idle - but I would expect it to drop the voltage in this situation, not leave it the same.

I've now overclocked from 2.4Ghz to 3.0Ghz, switched off several "Auto" voltage options in the BIOS and brought the CPU voltage down to 1.3V in the BIOS which is being reported back as an actual voltage of 1.26V by speed fan/cpu-z. I've also managed to undervolt the RAM down to 1.8V whilst keeping the 400Mhz 5-6-6-18 timings and now those temperatures are: CPU 47, Core1 54, Core2 56, System 45. I'm a long way from finished playing with it, but I suspect the large power draw those doing reviews are finding may be in part due to the board by default using increased voltages.

I've got no way of directly measuring it but the difference in heat should give a clue to the difference in power consumption from sensible BIOS settings - after all I'm now running 25% faster and 16% cooler than the default settings.

Also, there is the small matter of did they switch of the Wi-Fi transmitter?