AMD fries, Intel survives
 

Just finished watching a video published by tomshardware.com... Interesting. It featured four CPUs, coolers being removed, and results documented...

Intel Pentium 4/2000, Socket 423, Intel 850 Chipset
Temp 29C/84F
Application slows down, CPU survives

Intel Pentium III/1000, Socket 370, Intel 815 EP Chipset
Temp 38C/100F
Application freezes, CPU survives

AMD Athlon 1400, Socket 462, VIA KT133A Chipset
Temp 370C/698F
Application crashes, CPU & Board up in smoke

AMD Palomino 1200, Socket 462, VIA KT266 Chipset
Temp 298C/568F
Application crashes, Thermal diode fails, CPU up in smoke


You get what you pay for, I guess.

Why do you suppose this is?

the more pointed question is, why do we care? What kind of idiot would remove the heatsink from a running processor?

I sure wouldn't, but if the cooler malfunctioned, I'd prefer not to have to purchase a new CPU.

Unfortunately, you dont always get what you paid for, especially with intel. You pay more, and it seems often get less performance.

Yes, intel is more heat freindly - but who cares? Get a decent heatsink and fan on your chip and it doesn't really matter, does it? Also, at the price of an intel (especially back then), you can afford to go AMD, roast it, and buy a new one almost.

Yeah...and a new motherboard, too?

Anyway... I was asking about the technical difference. In case that somehow slipped between the cracks.

Just like those motor oil commercials that run the engines without oil, then make the jump that somehow thier oil is better. Who runs an engine without oil??

ummm.... based on the temps you posted, it looks like the AMD is surviving until almost 10x the temperature.... i would imagine that the intel CPUs would do the same at 698F.

kinda unfair comparison?

This has been discussed before (but it wasn't in it's own topic) here http://www.tfproject.org/tfp/showthr...threadid=45052

the relevant part of the discussion starts about half way down the thread.

Short version was that because stuff like this isn't "reasonable" (i.e. ripping out a heatsink while running) and that he newer processors have more/better (?) protection against this (both intel and amd).

Also to throw some fire into this discussion...
Intel processor burning pc
http://www.sudhian.com/showdocs.cfm?aid=494

(note that in both your case and the article i mentioned above, you can do stuff that was never expected to cause stuf to burn).

There is a concept that here that goes without saying.

The idea of being shown an engine running without oil is to say that if something happens to your car, your mechanic forgets to tighten your drain plug, your filter falls off or you blow an oil cooler line, you have a cushion.
By the time that you are made aware that you are out of oil, you engine will NOT be toasted.

The same could be said for your computer, someone walks into your computer room, accidently knocks over your tower, sets it back upright. You enter later that day and turn on the computer, and you were not aware that the heatsink fell off when the computer was knocked over. Within seconds the AMD chip is fried. If you had an intel chip, the system would be unbearably slow, and you would eventually discover that the sink/cooler fell off.

You can also look at it this way, AMD does not put a thermal shut-off in their processors because they want you to have to buy another one if it gets fried. You pay more for the Intel chip because they are not going through the volume that AMD does. Personally, I am all about having a backup device that prevents the chip from getting fried. If I buy an expensive chip, I sure as hell don't want to have to go through the downtime and replacement costs when I could have just paid more in the begining and not have to be hasseled.

Just my $0.02

-SF

Or you could simply get a mobo that has a heat sensor under the proc and an auto turn off at whatever temp is +5/+10 your underload temp.

Edit: Now that I think about it I just realized that an AMD socket A heatsink is IMPOSSIBLE to accidently knockoff. The three prong snap only comes off with a lot of force, a screwdriver and a good bit of luck.

Those are both old Athlons, and this is old, TIRED news. The videos that Tom's produced were partly responsible for AMD's implementation of better thermal overload control.

Is there any other oblsolete news you need to bring to our attention?

Bingo. That's why the article is so pointless. Unless you pull the heatsink on purpose, the heatsink isn't coming off and the chip won't fry. The only way this "problem" could present itself is if some asshole is out to get you and removes your heatsink. But then, wouldn't it be easier for the aforementioned asshole to just open your case and jam a screwdriver into the motherboard?

Can't cook no soup on an intel. :cool:

the A7V333r in my main system has a AMD 1900+ CPU and "ASUS C.O.P. (CPU Overheating Protection) for Athlon™XP only" (had to visit the website for exact phrasing). Never have had reason to see it in action but that should avoid the unpleasant results of that old but still farily relevant video.

Yeah, the very, very first Athlon XP's had this problem. After the 1.4 GHz models came out it was fixed, however. Either way, the new Athlon XP's are fixed. That video had relevance when those CPU's were current, but now it's simply useless.

Even then, the heatsink fan actually coming off is nearly impossible. Even if someone were to beat the motherboard hard or knock the computer off, the damned heatsink fan would rip the entire socket off the motherboard. If you've ever installed an Athlon XP heatsink fan, you know it's NOT coming off without the socket attached unless you get it off yourself. The only way this is possible is if the heatsink fan is installed incorrectly, and this happens more often than you'd think.

-Lasereth

On the cooler fan failing, the heatsink is still intact so heat is still transferred from the CPU to the heatsink. If you had some bloody good aeroflow in your case, it would probably function the same.

this video has been around for quite a while now. i would suspect that amd has done a lot towards fixing this problem, although i haven't seen any proof that they have. but with or without thermal protection, the hsf doesn't come off easily.

oh, i've heard not so great things about the thermal diodes on the boards. people claim that it still doesn't work when it hits the temperature you set.

and redarrow, you'd need some seriously strong airflow to keep it functioning the same. normal case fans that are so far from the actual hsf won't keep the proc cool enough.

merlin:

the problem with the thermal diodes on the motherboard is that depending on where the sensor is taking the temperature reading you will get wildly different temps. (i..e touching the bottom of the chip, in the socket, on top of the chip, etc). This is the reason why those mobo diodes don't work well.

hence the reason I used 'fairly relevant'. The only reason I am still not using a 1300 Athlon CPU was my uncle upgrading and giving me his castoff 1900 XP.

Everytime sombody brings me this news forward attempting to use it for a valid argument against AMD it only causes me to laugh at them, pass it off, and give that person less credibility.

Why even bother bringing up such old old old old news?

Back when this was made, it had *some* ability to be used in a argument, but even then it was hard to pull off. Now.. dont even waste your breath.

*sigh*

Is it possible to discuss something without it getting personal or completely off-topic?

The fact that this is old news to you is news to me as I had never seen/heard of it before. I never said it was brand spanking new, only that I had just seen it.

I only asked what the technical difference is/was between Intel and AMD in this respect. I don't give a fuck about your personal feelings regarding Intel, AMD, or the Theory of Relativity. Is this really that complicated?

Hey, you asked why this was happening, and we explained it. This video is old as crap and AMD has fixed their later Athlon XP's. :)

-Lasereth

I was surprised how much force it took to get the heatsink in place when I was putting my computer together. There's no way in hell that thing just "falls" off.

Easy. Intel has some thingamaboby that detected the heat buildup (or sideeffects thereof) and downclocked the CPU to keep it running.

AMD either lacked the thingamaboby or it wasn't working right, so it kept on going at full speed generating heat until it literally burnt itself out.

As noted, working anti-heat thingamabobies are now availiable on basically all modern processors.

Do you want me to confirm this and/or look up the marketspeak / engeneerspeak / slang for thingamaboby?

If your motherboard doesn't suck, it should kill power to itself at least, and others do the chip too.

Special Features (in table)



Well... for reasons I'd rather not ... get into right now ... I have seen it in action. When it senses your chip overheating, it kills all system power right away.

The only thing I have had happen to fry a chip was that teh CPU fan died, causing the heatsink to get extremely hot and melt the socket holding it in place. This allowed it to fall out of position and the CPU got a little bit hotter than is condusive to operation. I hate the smellof something burning coming from a computer.