Worthless article on frontpage

[Hezik]

About this article:

http://www.tomshardware.com/cpu/20041114/index.html

I am really disappointed in Toms Hardware bringing out such a badly performed test. The results of the test are clearly influenced by the unprofessional way of working.

First of, Intel DOES NOT recommend 38 degrees case temperature. Intel specifies 38 degrees as the max ‘normal’ working temperature. It doesn’t say ‘keep the case at 38 degrees’, it sais ‘keep it at or below 38 degrees’.

Second off, a cooler isn’t mounted with the case/motherboard standing upright. This is also not recommended by intel.

And then my largest point of criticism.. where the hell did you guys get the idea that such a gross amount of cooling paste is required? The amount you put on there is about enough for like 10 CPU’s or more.

For anyone who has any knowledge about CPU’s it’s obviously that this test is a fraud.

Don’t believe me? Then please repeat the test, but now:

1) follow the instructions of the paste supplier
2) use paste designed for usage with CPU’s
3) mount the CPU in the correct way
4) Clean both the CPU and heatsink before testing

Also you might consider using a non-boxed cooler for your test.

http://www.tweakers.net/ext/f/45176/full.jpg

<P ID="edit"><FONT SIZE=-1><EM>Edited by hezik on 11/15/04 03:07 PM.</EM></FONT></P>

 

[DonnieDarko]

wow stop crying, that article was meant to prove a point about the heat related problems intels chips have been having, which in fact is true.

You never actually see them put the hsf on. If you reread the box it does say they recommend a case temp of 38 degress or below

Intel recommends that you use a chassis with an internal ambient temperature at or below 38 degrees

Notice the <b>at</b> 38 degrees, meaning 38 is exceptable, which it isnt. With the thermal paste, yes they did put alot of but i think it was to show them applying it, thg knows how much thermal paste too apply. And if you read the whole article at the end they said how they cleaned the cpu and hsf off before adding the new paste.
quit being a fanboy just because they said something bad about a company you perfer. This was obviously a poke at intel, but in a way to show people the truth, there are many people that come on to these boards about thermal problems with prescots. And if you read the article they say that this was meant to show what it would be like for the average user, not a novice.

Watch out for the <b><font color=red>bloody</font color=red></b> Fanboys!

AMD64 2800+ :: MSI Neo-Fis2r :: 1024mb Kingmax ddr400 :: Sapphire 9800pro 128mb :: 10K WD Raptor

Addicted, finally.

 

[dkrenshaw]

Ok, I don't understand whats up with this article either. They bash on Intel for not providing an adaquate cooler in their boxed product... yet in the article on Nov. 3, they praise the Intel boxed cooler...

"Summary: You're Well-served With The Boxed Cooler....
...The boxed cooler from Intel achieves the best score when the heat-conducting paste supplied with the cooler is used."

http://www.tomshardware.com/cpu/20041103/cooler-09.html

and, in that same article, they claim that the high quality thermal paste is supplied with the boxed cooler!!!

"The cooler reaches just 0.37 K/W with the heat-conducting paste supplied with it, which is applied to the cooler in a similar way to a heat-conducting pad. When conventional heat-conducting paste is used, the thermal resistance rises to 0.51 K/W! For this reason, we recommend that you use the accompanying heat-conducting paste. As noted above, the cooler is included in the items supplied with the "boxed" version Pentium 4 CPUs. It cannot be bought separately."

http://www.tomshardware.com/cpu/20041103/cooler-02.html#intel_boxed_cooler


whats the deal???

 

[Vapor]

I really see little wrong with the article or the way the test was conducted.
1)

First of, Intel DOES NOT recommend 38 degrees case temperature. Intel specifies 38 degrees as the max ‘normal’ working temperature. It doesn’t say ‘keep the case at 38 degrees’, it sais ‘keep it at or below 38 degrees’.

No [-peep-], Sherlock. However, they didn't turn on a heater to warm the case to that level, the processor itself made it that high! They did *not* force an ambient temp of 38C!

2)

Second off, a cooler isn’t mounted with the case/motherboard standing upright. This is also not recommended by intel.

That's to prevent user error, as long as the HSF is 100% fastened and secured, it's fine (no matter how you do it). I'm 100% sure that THG fastened 100%, as seen in the video. Another reason why I know it's fastened correctly? EVERYONE else has major heat problems with their 560s!

3)

And then my largest point of criticism.. where the hell did you guys get the idea that such a gross amount of cooling paste is required? The amount you put on there is about enough for like 10 CPU’s or more.

I agree that the paste application was egregious, however, it does spread out when the HSF is applied, although not necessarily to the ideal thicknes. In addition, not everyone will apply the proper amount; in fact, most think that 'more is better.'

Even if you (wrongly) completely discount the thermal paste tests, the stock thermal pad tests are completely valid! It has also been shown a few times at very places that Intel's new TIM is better than generic paste and almost as good as AS5 (but not reusable).

As for your other points:
1) tons of people don't follow the directions for paste (as I said, more is better in a lot of people's minds), if there are any directions; somewhat valid point, yet if you go by what you said (since that paste doesn't have directions), they shouldn't have put any paste on at all!

2) that paste was designed for CPUs; moot point.

3) the CPU and the HSF, although not necessarily mounted properly (the process, at least) were mounted properly (end result, at least)--also, we don't know if the case was vertical or not at the time, we have no external reference in the video; COMPLETELY moot point.

and 4) can't clean the HSF before testing with the included TIM and the CPU was adequatly cleaned for the second (and first) two shots--the HSF's cleanliness in the second shot is still unknown (though I did do a personal test a few months ago--kept the stock TIM on the heatsink, cleaned the processor and added some generic silicon paste--I did not observe any different temps than what the generic paste did on its own! [and this TIM is better than the old stock one!]). Would be a valid point if it made a difference in the end results.

BTW, the <b>POINT</b> of the article was to test the parts that Intel gives you!! What good is a product if it doesn't work right w/o other parts?!

Maxtor disgraces the six letters that make Matrox.

 

[Vapor]

whats the deal???

It maybe the best cooler, but it still isn't good enough! This is Intel's fault on two fronts: creating a processor that dissipates so much heat and not making a cooler that is good enough (as good as it may be).

As for the included TIM (which is new), it has been proven a few times that it is better than generic paste (and about the same as AS5).

To note is that their initial tests are done in an open-air environment while this most recent test was done in a closed case in order to show the real performance of the HSF.

Maxtor disgraces the six letters that make Matrox.

 

[Hezik]

No [-peep-], Sherlock. However, they didn't turn on a heater to warm the case to that level, the processor itself made it that high! They did *not* force an ambient temp of 38C!

In the movie that comes with the article, they clearly state that 38 degrees is the recommended temerature. Also, they show their case to be 38 degrees. Though 38 degrees may be intels specified max, it's by no means a temperature one should strife for. Indeed they didn't force this temperature, but they did nothing to couter it either. If your case is 38 degrees on the inside, there's something seriously wrong with the airflow.

I agree that the paste application was egregious, however, it does spread out when the HSF is applied, although not necessarily to the ideal thicknes. In addition, not everyone will apply the proper amount; in fact, most think that 'more is better.'

The point is that ideally there is NO paste between CPU and Cooler, only metal-to-metal contact. Paste is only ment to fill up impurities, scratches, and so on.

Also the screenshot i posted clearly shows the CPU was not cleaned properly before applying the paste. If you watch the video closely you also see that this paste was NOT ment for CPU's, it's an industrial size tube used for heatsinks on things like mosFETS etc. Wichever way you put it, it's no arctic silver, or any other well know brand CPU paste.

If one wants to match an intel cooling solution to a 3rd party cooling product in all fairness, one should mount both properly. Clearly that didn't happen in this test. Seems to me this test only has one purpose, and that's making propaganda for intel stock coolers. This is also confirmed by the words 'buy intel coolers AND motherboards' in the movie.

They didn't prove anything about an intel motherboard, nor about a 3rd party cooling solution. If this test were to prove anything at all, then its that an Intel Thermal pad outperforms a gross amount of cooling paste.

That I could have told them in advance.


btw, this is the video i'm referring to:
http://www12.tomshardware.com/images/thg_video_14_intel_heat.zip

and this is the conclusion:

CPU is at it's limit
Intel PIV 560 (3600Mhz only) delivers best performance with boxed cooler

Avoid 3rd party coolers and motherboards

Now please tell me how ANYTHING in this video proves that 3rd party coolers or motherboards are bad?

Pure intel propaganda.. how much did they pay for this commercial?

<P ID="edit"><FONT SIZE=-1><EM>Edited by hezik on 11/15/04 05:47 PM.</EM></FONT></P>

 

[DonnieDarko]

every page I go to on this website has an intel ad on it, so I dont think this was an ad for amd. Do you have a p4 560 and know how hot they run? I have seen many people with throttling problems with this chip. I dont see why you are so upset about an article unless you work for intel. fanboys are so sad.

Watch out for the <b><font color=red>bloody</font color=red></b> Fanboys!

AMD64 2800+ :: MSI Neo-Fis2r :: 1024mb Kingmax ddr400 :: Sapphire 9800pro 128mb :: 10K WD Raptor

Addicted, finally.

 

[Hezik]

Maybe you should read what i'm posting. No i don't work for intel and I'm indifferent bout the intel vs AMD issue. I have both at home, both perfectly fine CPU's.

What i'm 'upset' about (not really) is the way this test is performed. Especially the video. It has nothing to do with testing stuff, but everything with commerce.

 

[jammydodger]

If your case is 38 degrees on the inside, there's something seriously wrong with the airflow.

...or you are using a pentium4 560

<P ID="edit"><FONT SIZE=-1><EM>Edited by jammydodger on 11/16/04 00:57 AM.</EM></FONT></P>

 

[jihiggs]

i think your over reacting, and you seriously look like a fan boy.

wpdclan.com counter-strike game server - 66.150.155.52:27015
now featuring valve security module!

 

[Hezik]

You need reading glasses..

I'm not saying anything pro or contra intel. If anything it's more anti-Intel. I have nothing against Intel or AMD, they both make fine CPU's, though Intel DOES have problems heat-wise.

I'm not commenting on wether or not intel CPU's get to hot or anything. My comment is about how this test is performed.

If you don't see that it's highly biased then YOU're the fanboy, a tomshardware fanboy to be exact. Nothing wrong with that, those people can be perfectly happy as well.

 

[Ned_Flanders]

this thread sucks.

Thread should be locked, and threadstarted should be banned / killed / raped (in that order)

_______________________
<A HREF="http://www.moviewavs.com/MP3S/TV_Shows/Simpsons/flanderssong.mp3" target="_new">Audio Sig</A>

 

[jihiggs]

oh now i get it, my opinion differs from yours so i cant possibly be right, my whole life up to this point has been a lie!

wpdclan.com counter-strike game server - 66.150.155.52:27015
now featuring valve security module!

 

[sobelizard]

If you send me an ATI 9800 Pro 128, I'll be your fanboy.

<font color=green><b><i>Lizards</font color=green></b> do <b>not</b> taste like <b><font color=yellow>chicken</font color=yellow></b>,<b> <font color=yellow>chicken</font color=yellow></b> tastes like <font color=green><b>lizard.</b></font color=green></i>

 

[jihiggs]

deal!





...sucker!

wpdclan.com counter-strike game server - 66.150.155.52:27015
now featuring valve security module!

 

[Hezik]

bah.. calling me a fanboy while you got your head up tom's .. anyways.. never mind..