Sign in with
Sign up | Sign in
Your question

amd dies, intel survives: other factors.

Tags:
Last response: in CPUs
Share
September 18, 2001 5:14:51 AM

interesting article that,

despite all the very predictable flaming going on, of which i want no part, there were a few interesting bits in there.

namely... if the pentium4 is such a prestigious heat outputter, how come its so cool? is the throttling that good?

also very interesting was the
bit with the palomino and the mobo only detecting a 1deg/sec rise, with the drastic rise being too quick to prevent switchoff. i wonder what plans amd has to change that? any ideas if its comming with the 0.13 micron core?

as for the comment about "check your heatsink once a month" thats a bit anal-retentive aint it? i havnt checked mine once yet and its been on two trips out the house.

final thought which the completely missed (shame)
SCENARIO:
you have a tower/mini tower case. pretty popular nowerdays.
mobo is vertical, with heatsink extending at right angles.
what do u think whould happen if some 300 odd grams of aluminium/copper suddenly broke free?
my guess it would accererate downwards at 9.8m/s/s and smash into the back of the graphics card with significant force.
so regardless of which processor you have you may be up for a new graphics card too!



In memory of the 90+ Auzzies missing in the WTC disaster. An attack that has changed the world.
Anonymous
a b à CPUs
September 18, 2001 7:51:26 AM

>what do u think whould happen if some 300 odd grams of
>aluminium/copper suddenly broke free?
>my guess it would accererate downwards at 9.8m/s/s and
>smash into the back of the graphics card with significant
>force.

Indeed. I think this whole thing is a non issue really.
Maybe Tom should now test how long a cpu lasts if you smash it with a hammer, or poor coffee on it. Maybe intel can come up with an integrated airbag that activates when the computer is dropped from 10 foot high.

Everyone already knew this. Take off your hsf and your cpu is dead. I admit its quite astonishing the P4 survives this and keeps running, but I couldnt care less for a home built pc. Maybe for a rackmount server farm in an earthquake sensitive location this might help though..

---- Owner of the only Dell computer with an AMD chip
September 18, 2001 8:19:36 AM

Yeah, the whole point of that article escapes me.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
Related resources
Anonymous
a b à CPUs
September 18, 2001 9:00:39 AM

Just means Intels CPU's are more idiot proof than AMD's CPU's. But for most of us you want to overclock AMD is the way to go and we are aware of it, best part of AMD is we can enjoy freedom in this respect..


<font color=purple>Three ways to do things, your way, my way and the right way!</font color=purple>
September 18, 2001 9:28:52 AM

probably just trying to get AMD to do something about it.
also it is nice that FUGGER and his mob have some hard evidence for a change.

Although it has a lot of good ideas, beer doesn't know anything about computers!!!
September 18, 2001 9:50:48 AM

We can see where you are dumb then dont we. It takes a lose wire on the fan.
Anonymous
a b à CPUs
September 18, 2001 12:03:29 PM

Loose wire, MBM5 would have plenty of time to shutdown by the time the HS got hot enough.


<font color=purple>Three ways to do things, your way, my way and the right way!</font color=purple>
Anonymous
a b à CPUs
September 18, 2001 12:06:28 PM

The possibility of falling off the H/S might be rare but still I would be happy to see some sort of heatprotection on our AMDs. I would be less nervous after playing around with H/S and compound trials. At least at the level of P3s. Putting down a thermodiode shouldn't be that difficult.


<font color=red>Apple</font color=red> <font color=blue>inside</font color=blue>, <font color=green>idiot</font color=green> <font color=purple>outside</font color=purple>...
Anonymous
a b à CPUs
September 18, 2001 1:06:24 PM

As with others, I have a question of the validity of this test. Tom indicates that it is common to have a fan failure, but performs the test on the worst case senerio (complete removal of the heatsink while operating the computer). I think that this review of the CPU's would of been better if the following tests were performed prior to the worst case senerio that we were given:
Perform two tests of each of the following conditions on a "cold system" - a system at ambient temperature and on a "hot system" - a system that has been running at full load and has reached a constant temperature

Do the following tests first on a "cold system" and later on a "hot system":
Test #1 - Heat sink fan stop working - run the system with the fan disabled or blocked
Test #2 - Heat sink removed

The problem with the original article does not represent the circumstance that Tom is originally indicating may happen during transport of your computer. If this test was performed on a "cold system" with the same results, I don't think there would be as much flaming going around, IMHO.
Anonymous
a b à CPUs
September 18, 2001 1:31:34 PM

THG was showcasing this new way of reviews for the future, so a bit of grandstanding is acceptable.
Flames are fun and brings the community together as well as purist opinions from some very knowledgeable operators..

Damn good clip btw.



<font color=purple>Three ways to do things, your way, my way and the right way!</font color=purple>
September 18, 2001 2:03:07 PM

I don't know how to feel about Tom's article... I thought we all knew that 1.4Ghz T-bird will die if you run it without heatsink anyway. It would be cool if Tom would test video cards in same manner: cut heatsink away from GF3 and film how it burns -> conclusion would be that nvidia makes interferior video chips :)  Oh, and even thought P4 would survive without heatsink, does video card that takes a hit from at least 500g heatsink (Intel spec if I remember right) would always survive, too? :) 
September 18, 2001 4:09:11 PM

hehehe.. have you seen the size of the HSF compared
to the 0.13micron P4??? Wow.. if that thing ever
comes off its taking a chunk of your mobo with it
:lol: 

Intel Components, AMD Components... all made in Taiwan!
September 18, 2001 4:42:40 PM

I have seen claims of where the heavy duty HSF's are blowing out the mother board power connector.

"Yeah, the whole point of that article escapes me" - are you sure you work in a fab?

Not sure why you got it lodged in your thick skull that the HSF has to "fall" off. there are too many ways possible for a HSF to fail. if that HSF is your only means of air flow around your CPU its gonna fry itself the like thousands that alread foound themselves on RMA and hanging from keychains..
September 18, 2001 5:10:08 PM

lol! you have a good point. The article i think is just there to show AMD how inferior their thermal diode really is. But your right take the heatsink and fan off a geforce and it would probably get hotter then the AMD chip lol ... i bet it would actually flame up lol.

it's like nvidia and AMD vs Intel and Ati ... weird huh... i like my ati however .. i don't get burned when i go change the IDE cables around lol ... i swear the TNT card got soo HOT! I can't imagine how hot those geforce 2 gets. Of course this was a couple years ago and yes i'm constantly in my computer playin around with it.

Next step ATI AIW Radeon 8500DV to go please! :) 

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
September 18, 2001 5:30:15 PM

"<font color=blue>As with others, I have a question of the validity of this test"</font color=blue>

The test was valid. It checked for the effectiveness of the thermal protection in each CPU, under the worst case circumstances. I don't think anyone was surprised by the burning of the AMD chips, even the Palomino.

For myself, I was impressed that the P4 was not damaged, and continued to run. I do think that Tom screwed up somewhere, because the chip had to be much hotter then 29c if it was throttling. It seems logical to expect the temp should have been around 70c, where the thermal protection kicks in.

The reality of it is how often does a HSF fall off? I have never seen one fall off. How often does a ball bearing HSF fail? I have never seen one fail. The chances of HSF failure are very slim. But seeing the thermal protection at work was a great article, and Intel deserves praise for its effective thermal protection.

<font color=blue>This is a Forum, not a playground. Treat it with Respect.</font color=blue>
September 18, 2001 5:56:46 PM

AMD's run hot. proven. Intels run cooler. proven.
AMD's run better. [proven]. Intels don't. [uuhh]

Frankly, although i'm worried about the heat problem, i don't have a "heavy" heatsink, and i'd rather have the bang per buck,

I'm perfectly satisfied with my Athlon 1400, it beats all my friends 'COOL' P4's so i'm happy?!

---

though AMD should follow Intels lead and integrate better thermal protection...

"The answer to life's problems aren't at the bottom of a beer bottle, they're on TV."
Anonymous
a b à CPUs
September 18, 2001 7:48:32 PM

The reasons why I question the validity of this test is that it does not represent what would occur in a machine except in the extremely rare, if not impossible, case when a computer was running full load and had the heat sink fall off of the cpu.

Like I stated in my previous post, if Tom had removed the heat sink on a cold system and booted up; or messed around with the fan, then we probably would see things that are more likely to occur.

I'll admit it is impressive that the Intel chips did not immediately burn themselves up when the heat sink was removed, but I just don't think that this test is representative on what could normally occur to test the thermal diodes.

I also have a slight problem that this article was based only on one test (I know, how many chips need to burn up until your're convinced), but it may have been a fluke for the Palomino or the P3. Also, as some others have wondered, was the Palomino a failure of the diode, or was the chip damaged when the heat sink was removed. If you look at the video, all of the other CPU's except the Palomino were still showing Quake when and after the heat sink were removed; whereas the Palomino's screen was black even before the heat sink was totally removed (time stamp 1:43).

Nobody was surprised when the Thunderbird fried, but the Palomino is a different story. If the purpose of this story was to show the effectiveness of the thermal protection on each type of CPU, then there should have been multiple tests performed of different conditions that could occur, with the last test being the one Tom performed. Finally, the data should be revied before making the final decision on what really caused the failure of the chip was not caused by something not being evaluated (e.g., damaging the CPU while removing the heatsink).

I know many people like reading Tom's articles; but when you see things like this, you begin to wonder if Tom is being objective against AMD products. IMHO I think it would be best if he did additional testing on both Intel and AMD chips to get a better idea on the effectiveness of the thermal protection on each type of CPU.
Anonymous
a b à CPUs
September 18, 2001 8:14:35 PM

In response to your question, Dave apparently was using an infrared thermometer from DataTRAK that, if my memory is correct, measures temperature based on infrared emissions.

Keep in mind that only the "outside housing" of the chip was being measured and *NOT* the inner-core of the chip. Generally there is a difference between these two temperatures (with the core running hotter, of course). I would expect the core to be around 65C. +/-5C.
September 18, 2001 10:11:40 PM

I would like to point out a real life situation that my Tbird 1400 survived. I was checking airflow in my system (it gets hot in there with a 1400 and several 10000rpm drives) and I ended up pushing the power cable to the fan header, and the fan stopped. Now I know, dont play with a running computer (do it all the time) but the fan stopped for a good 5 seconds before I pulled the power cord. After reattaching the fan power it booted right back up. I dont know how relevent this is, but it does show that the chip can survive for a short period without the fan running =)

Jesus saves, but Mario scores!!!
September 18, 2001 10:15:24 PM

It's a lot more relevant than Tom's article, if you listen to some people.
It's been proven before that a Tbird can survive with only a passive heatsink for a little while, and MBM (or whatever) can save your chip.

With NO heatsink, it's a different story.

Hmm...your sig reminds me of one I saw that was "Jesus saves, and takes 2 damage.", or something like that.

<font color=green>In memory of all the Americans that died 9/11/01
Rest in peace</font color=green>
September 19, 2001 12:58:19 AM

nice
yes... thats another valid thing tom DIDN'T do

he should have also tested what happened if something more realistic happened,
e.g. fan failure or dissconnection

the reason why your 1.4 didnt fry is that the heatsink still works without out a fan, just alot less efficiently. the block and fins still distribute the heat over the entire heatsink, but only has natural convection currents and heat diffusion to the air to cool it.

i suspect in this test 3 of the 4 processors will have survived. the temp climb for the palimino being slow enough for the mobo to shut of.

come to think of it... with my asus a7v133 & 1.2C, if the hardware sensor for the chip ever tops 75C the mobo has orders to shut down. so im protected from fan failure in this case.

i still shudder to think what a falling HSF would do to the internals of a case. graphics card damage. mobo damage. scrapes on the surface cutting and kioning electrical paths. capacitors knocked off. yum.

on a more positive note, my half hour with some wirecutters on the back of my case has yielded a nice gaping 80mm hole which my casefan can now effectivly blow though. 3-4C off the mobo temp, and 2-3C off the processor temp.
cheapest cooling solution ive ever seen *grin*

In memory of the 90+ Auzzies missing in the WTC disaster. An attack that has changed the world.<P ID="edit"><FONT SIZE=-1><EM>Edited by lhgpoobaa on 09/18/01 09:01 PM.</EM></FONT></P>
September 19, 2001 1:15:22 AM

damn you all, i'm getting a desktop case now.
September 19, 2001 1:56:17 AM

relax!
we ARE talking about absolute worst case scenario here.

desktop cases do have the advantages like more space inside, possibly better airflow, and a verical mobo that doest gather dust unlike a desktop.

methinks the chance of your clip breaking is less likely than getting a lightning strike that frys your pc.

In memory of the 90+ Auzzies missing in the WTC disaster. An attack that has changed the world.
September 19, 2001 2:01:18 AM

Relax, i'm kidding. I'd beg to differ on the space issue though. All the computers i have owned have been desktop cases...and absolute bitches to mess around with. My friends all seem to have midi towers. God they are so much easier when it comes to adding/removing components. Also, would you recommend leaving off the side panels of a midi tower to keep it cooler?
September 19, 2001 2:22:45 AM

No add some case fans and put the sides back on.

Nice <b><font color=green>Lizards</b></font color=green> <b>crunch</b> Trolls cookies....... :smile: Yummy!! :smile:
September 19, 2001 3:32:10 AM

I feel special- my HSF and CPU slot are next to the vid card, so if it falls, it'll fall on teh bottom of my case. This way no resulting damage to other parts. I really liked the clip; it looked and sounded cool. If u want smoke btw try a P1 w/ an incorreectly modified mobo so like 20x electricity goes to processor. The thing burns, as does the whole mobo! Fills teh whole room up w/ smoke. Anyway, I personally think I'd be happy if my HS fell off; that would mean that I would RMA my 1.2, pay a few bucks more, and get a 1.4 :)  I don't think it'll happen, as I never carry my computer around. Laptops are meant to be carried around, not big PC's. In addition, the GF3 will not catch on fire, hell mine doesn't even get very hot after I play on teh comp non stop the whole day and touch it while it's running. I think thermal issues should never stop u from buying a processor. Use liquid nitrogen if u have to.

Your brain: PC
You brain on drugs: Mac
September 19, 2001 3:46:07 AM

I have a Desktop Case. So i don't have to worry about my graphics card.

Nice Nvidia and ATi users get a Cookie.... :smile: Yummy :smile:
September 19, 2001 4:24:14 AM

sure u can run it with the sides off!
same as u can drive a car with the doors open if u want

it will be cooler, but will also be noisier, gather more dust and have all the inherent dangers of an open case (things falling in it, pets sniffing it etc)
just get a couple of casefans and ensure you have adequate air circulation.

gee i just love the way i savaged the rear casefan hole :) 


In memory of the 90+ Auzzies missing in the WTC disaster. An attack that has changed the world.
September 19, 2001 2:25:14 PM

FUGGER if a fan dies on a heatsink the CHIP WILL NOT FRY WITHIN 1 MINUTE, motherboard monitor can shut it down with plenty of time to spare, there are tons of posts on this board confirming just that.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
September 19, 2001 2:28:15 PM

Quote:
Hmm...your sig reminds me of one I saw that was "Jesus saves, and takes 2 damage.", or something like that.


LoL I love d&d humor.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
September 19, 2001 2:34:28 PM

are you talking to me?

Intel Components, AMD Components... all made in Taiwan!
September 19, 2001 9:13:48 PM

I kinda like having the sides off my case. When we put sheetrock up in my room we made a hole for computer cords to go through, so I have the monitor and KB inside and the case outside. Quiet and I can still have a 22 inch box fan pointed into my computer. Cools it 10 to 20 °F or more depending on room temp. I figure this winter when the outer room is below freezing I can try some awsome overclocking.

The most useless computer tasks are the most fun to do.
September 19, 2001 9:59:24 PM

I think that it mearly served to highlight that thermal output is become a serious problem for processor architects. It demonstrates how the two companies are dealing with the problem in different ways, one, by the looks of things, not at all, the other is making an attempt although that is probably reflected in the pricing of the parts. But such things as the heat spreader seem to solve the problem of cracking the core if the heat sink is note placed with enough TLC. Its possible that someone might want to put a P4 in a rack mount with out sufficient room for a heat sink. Unlikley but possible.
But as I am being taught now, the idea of engineering to make a product "just" work. So perhaps INTC's products are a little over engineered.

At the end of the day "[-peep-] happens" and

Must go faster, must go faster,
Heck, processors fried...
September 20, 2001 1:20:39 AM

i think you raise a good point there Mr UK.

im curious as to what impact does the complex meltdown protection of the Pentium4 have with regards to its final shelf price?

also... i wonder how many UN-computer savvy people out there have P4's with a failed fan or incorrectly seated heatsink.
and just dont realise that the performance is being hobbled because they would never think to open the case?


In memory of the 90+ Auzzies missing in the WTC disaster. An attack that has changed the world.
September 23, 2001 1:25:14 AM

I loved the video to this article. I don't know how I missed it. I hope Tom continues the melting of cpu's and maybe GF3's as well. Still I am impressed with Intels cpu's having this sort of protection ability. A GPU such as the GF3 and Radeon 8500 have even more transitors then a P4 and probably just as much heat if not more. Maybe AMD systems and high end 3d card should come with a fire extinguisher on the side of them. :smile:
Anonymous
a b à CPUs
September 23, 2001 4:12:10 AM

would be good to know how Motorola's chips for the Macs cope with loss of their cooling devices.
September 23, 2001 4:15:56 AM

Who says more dust gathers when the sides are off your PC.. if anything it reminds you to blow a can of air in once a month or as needed, have you ever taken someones case apart after its been closed for a year or two? Leave the sides off. It's easier access for the kirby
September 23, 2001 4:33:08 AM

Personally, I loved the video. Ill never have to answer another post "What happens if I run my Athlon without a heatsink...Just for a minute?"

P.S. What music did they use in the film? With some tweaking, it could be quite a club mix.

"If you teach a child to read, then he or her will be able to pass a literacy test" - George W.
!