Why do reviews skimp important points??

[CAT-THE-FIFTH]

I was reading this review for the GA-P55M-UD2 and its overclocking ability:

http://www.bit-tech.net/hardware/mot...board-review/1

They were saying how the uncooled VRM would not be a problem for overclocking. However cooling the VRMs is more about long term stability as opposed to how far you can overclock!!

They should be at least advising people to use coolers that will pass some cooling air over them or having a case with decent airflow.

A member of their very own forums damaged the same motherboard as the VRMs gave up during stress testing:

http://forums.bit-tech.net/showthread.php?t=181448

Running a huge overclock(4.1GHZ on a Core i5 750) on such a motherboard with no sort of VRM cooling could cause issues in the long run.

[Pob255]

I think you ment Why not Who

and they did cover there collective asses with

The CPU socket has a basic six-phase power with no additional cooling on the MOSFETs, but that didn't prove a limitation in our air-cooled overclocking. For more extreme overclocking, you'd be better off spending more than £85 on a motherboard anyway.

Although they really should of added for long term overclocking in there as well.

However yes I do agree with you, saying they got these very good overclocks without really stressing the possible weaknesses is a bit slap dash.

[kalniel]

However cooling the VRMs is more about long term stability as opposed to how far you can overclock!!

No it's not. I've not seen anything to suggest cooling VRMs has an effect on long term stability. Electron migration is caused by overvolting, not heat.

[Zak33]

title edited away from my surname... Who, to Why

[Saracen]

On that particular product, I've no idea because I haven't read the review.

In general, though, points will be skipped for one of several reasons, including :-

• space limitations and more pressing points. Applies more to print than online, but still, if you have a wordcount to hit, you have a wordcount to hit, or
  
  
• the article is aimed at a different readership, for whom that point may not be relevant, or
  
  
• the article is aimed at a different readersip, for whom that point is to technical ... or not technical enough, or
  
  
• the reviewer simply missed it. It's a poor show, but happens, or
  
  
• the reviewer didn't spend the necessary time, because he makes more money if he simply does a job that gets past and moves on to the next review, or
  
  
• the review is a numpty that couldn't find his/her backside with both hands and a bloodhound.

In some 20 years and in writing several thousands of reviews, I've come most of the above, and have been in the first four situations.

[CAT-THE-FIFTH]

No it's not. I've not seen anything to suggest cooling VRMs has an effect on long term stability. Electron migration is caused by overvolting, not heat.

Plenty of evidence actually.

If you had bothered to read the review they have not overclocked on stock voltage and have actually upped the voltage to 1.35V which is only slightly under the maximum safe voltage of the processor(1.4V). Pushing more voltage though the processor means more strain on the VRM components and more heat production. The VRMs components are only rated for a certain temperature range as with all electronics components.

Plenty of components such as MOSFETs have a working temperature range and if you exceed this you will not have optimal performance and reduced lifespan. I have known plenty of people who work with electronics myself.

Even Gigabyte say that cooling is important.

For example look at the web page for the GA-MA770T-UD3P motherboard:

http://www.gigabyte.com.tw/Products/...ProductID=3096

They say the following:

"* Please apply sufficient cooling to CPU VRM zone when ACC function is enabled."

[kalniel]

Plenty of evidence actually.

If you had bothered to read the review they have not overclocked on stock voltage and have actually upped the voltage to 1.35V which is only slightly under the maximum safe voltage of the processor. Pushing more voltage though the processor means more strain on the VRM components and more heat production. The VRMs components are only rated for a certain temperature range as with all components.

Yup, so as long as they stay within that temperature range their longevity is not affected.

Plenty of components such as power transistors have a working temperature range and if you exceed this you will not have optimal performance and reduced lifespan. I have known plenty of people who work with electronics myself.

Even Gigabyte say that cooling is important.

For example look at the web page for the GA-MA770T-UD3P motherboard:

http://www.gigabyte.com.tw/Products/...ProductID=3096

They say the following:

"* Please apply sufficient cooling to CPU VRM zone when ACC function is enabled."

As above - you need cooling to stay within the tolerable range. But if it's stable within the range then the lifespan isn't going to be affected by being at the upper end of the range than say the lower.

[CAT-THE-FIFTH]

Yup, so as long as they stay within that temperature range their longevity is not affected.

What if they are not within the temperature range?? What about the fact that you need to consider the heat actually produced the component itself and the other components around it. How high is the local ambient temperature around the VRM area then especially if you are using a water or tower cooler??

Nice to also ignore the fact that Gigabyte has stated cooling is important.

[kalniel]

What if they are not within the temperature range??

Then that would be a point to mention if the review if it had any effect on stability. If it didn't go out of range or had no effect then I wouldn't think it was worth mentioning - you could write 100 pages of things that aren't problems.

What about the fact that you need to consider the heat actually produced the component itself and the other components around it. How high is the local ambient temperature around the VRM area then especially if you are using a water or tower cooler??

That's all taken into account - it still won't have an affect on longevity.

Nice to also ignore the fact that Gigabyte has stated cooling is important.

If gigabyte state that you might need additional cooling to stay within spec and stable then fine, but if a reviewer finds that the airflow in their setup is sufficient to cool it passively so that it is stable then I'm not sure what the problem is.

[scaryjim]

Then that would be a point to mention in the review if it had any effect on stability.

I think the real issue is that, unless the reviewers push a board significantly past its thermal limit and cause an instant catastrophic failure, there's very little chance of them knowing whether the VRM components are operating just within, or just outside, of their rated temperature specifications. Since reviews *have* to be turned around within a few days, they simply can't know the effects of running that same overclock for a month - running < 1 degree over rated spec for 24 hours might not cause an immediate problem, but running for a month just might. And it would only take twenty seconds to add "However, it would be worth adding extra cooling to the VRM areas if you plan on running your rig overclocked all the time."... It might get a little tired adding that to every review, but most readers won't read every review, they'll read a small percentage that appeal to them...

[Queelis]

So I'll just ask to be on the safe side should I ever overvolt (currently undervolted) - if I have this motherboard, I assume the heatsink with "Ultra Durable" written on it is cooling the VRM, so it should be OK?

[Terbinator]

I suppose there get out clause is "We take no responsibility for you following this guide... Overclocking is running components out of spec - they may break"

[scaryjim]

Queelis - yes, that heatsink does cool the VRMs, so it will certainly help. But, as always, there's the disclaimer that terb has mentioned above - overvolt / overclock, and you're running components outside of their rated tolerances, so there's always a chance something will go wrong!

Nice mobo, btw