Sign in with
Sign up | Sign in
Your question

AMD's response to Tom's Hardware CPU burn-up

Last response: in CPUs
Share
Anonymous
a b à CPUs
a b V Motherboard
October 26, 2001 2:17:23 AM

http://66.26.27.33/files/amdburn.zip

Seems like Tom just didn't use a motherboard that supports thermal diode for Athlon Xp/Mp Cpu!


Why didn't Tom mention about that ?
October 26, 2001 2:22:38 AM

And you see this how?

<font color=blue>Another waste of bandwidth on the web. :tongue: </font color=blue>
October 26, 2001 2:23:40 AM

Try rereading Tom's article. He clearly stated on the <A HREF="http://www6.tomshardware.com/cpu/01q3/010917/heatvideo-..." target="_new">second page</A> that he used a motherboard specifically designed to make use of the Palomino's thermal diode.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
Related resources
Anonymous
a b à CPUs
a b V Motherboard
October 26, 2001 2:28:15 AM

Have you checked that video ? it is only 2.6 meg

If that's the case, how come that Athlon never burnt even the heatsink was removed ?
October 26, 2001 2:36:52 AM

I started the download but it was way too slow. Then I realized it came from an unknown site without even a hostname. I do not download anything from unknown sites. Even if it really is a movie, since it is from an unknown (and hence untrustworthy) source, I would imagine they could have done any number of things to keep it from burning up. For example, you could have two computers, one running some benchmark while the other is off. Just set your camera on the one that is off and remove its heatsink. It certainly will not burn up, will it?

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =
Anonymous
a b à CPUs
a b V Motherboard
October 26, 2001 2:47:25 AM

I can understand your feeling for downloading stuff from an unknown site. Let's hope it will become available later in other "brand" name website thne you will understand when I meant in my very first post.

It is a video taken by AMD. It shows that HSF was taken off while when the Athlon was running Quake3. The computer just become blank without demaging the chip, and the Athlon keep running for more than 9 mins when the CPU fan was unplugged.
It is just very different to what Tom has shown us before.

This video is not super clear but it has some chessy music too.
October 26, 2001 2:55:32 AM

<i>"The next bit of news we have to report isn't as positive and it is in regards to support for the Athlon XP's on-die thermal diode. One of the mobile centric features of AMD's new Palomino core is the long awaited introduction of a thermal diode to more accurately measure the chip's temperature on the die itself. Unfortunately, as we saw when the first Intel processors began shipping with on-die thermal diodes years ago, motherboard manufacturers are very slow to support the diode with their designs. All three of the boards we evaluated continued to use external thermisters to measure CPU core temperature in spite of the presence of the new thermal diode in the Palomino cores. This was to be expected as it wasn't until the latter days of the Pentium III and really until the Pentium 4 that we saw motherboard manufacturers read CPU temperatures directly off of Intel's on-die diodes. While it would be nice to see a quicker adoption phase with Socket-A motherboards we're not getting our hopes up. As you can probably guess, a major feature for all of the current and upcoming KT266A motherboards is full support for AMD's Athlon XP processor."</i>
-Anandtech, VIA KT266A Initial Roundup; 10/24/2001


Now, this isn't Gospel truth or anything like that. I'm just saying that it's Anandtech's impression that there are no chipsets out there that yet support the on-board diode yet. Now, I realize that THG states that the KT266 has the hardware to use the on-board diode, but to have the KT266A not have it makes no sense. Therefore, someone, somewhere is wrong, and I couldn't tell you who. But, I have read from other sites as well that current chipsets (or motherboard makers) do not have the on-board diode reading capability. *Shrug* Maybe it's a matter of the motherboard manufactuers not implimenting the traces and circuitry to take advantage of it.

-SammyBoy
October 26, 2001 4:43:37 AM

Sammy is right, the older board would not have had the thermal hardware if the newer one does not, also just because tom says the mobo was equipped to handle the palomino does NOT mean that it was. Tom may have thought it was, but only tom is saying it was.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
October 26, 2001 5:05:32 AM

yeah but if it were an intel processor you would be crying bloody murder, but since it's an amd it's 'toms fault' that it started smoking. Ok sure there! :) 
October 26, 2001 5:08:07 AM

Dont presume to tell me what I would think troll.

I am saying that tom says the mobo was capable of using the palomino's thermal diode, and several other sources claim otherwise, only tom is claiming it could.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
October 26, 2001 5:24:21 AM

Intel_inside, some friendly words of advice for you.

When you are going to post a response to post, or an opinion, do try to be civilised about the matter, educated in the matter at hand, and do not suggest what others on the board may or may not say. Make sure you know what you are talking about before you post, and if you are not sure, but want to express an idea, do so politely: For example,

"As I understand it, isn't RDRam the better ram solution for the future, and wouldn't this be a problem for AMD in the future? And how exactly are the rates of transfer calculated for RAM, as I get confused with GHz and MBytes/Second."

...sounds better than:

"AMD is screwed, Intel is going to have RDRam at 6.4 Mbytes/Second and Atholns don't support it, this is 12 times faster than 266MHz DDR Ram, so Intel is still the best".

You see? Raystonn is probably the best example of a civilised poster: He leans towards Intel, which is fine, but he doesn't slag off AMD, he explaines his ideas logically, demonstrates that he has a firm grasp of the matter at hand, and doesn't use derogatory terms in his text.

As it applies to this post, Tom would have been using Intel motherboards that do use the internal diode of the Pentium, while the Athlons being tested did not have motherboards that used the diode to manage an overheating situation. This is why people believe that the Atholon melting tests may have been negatively biased when it comes to the issue of overheating. Suggesting that someone would have 'screamed bloody murder' is immature at best. If other people have an opinion, they will express it, hopefully in a civilised and respectable manner as well.

Cheers,

Ozzie

*


How long is a piece of string? The same shape as something beige.
October 26, 2001 7:46:58 AM

That guy intel inside is just a loser here to start trouble, I say everyone should ignore him from now on, or spam his email to those of you who did that ;-)

::applauds::

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
October 26, 2001 7:54:29 AM

"And Intel_inside tumbled as he was attacked by a group of 100 12-year olds whipping spitballs and spewing verbal slander..."
October 26, 2001 8:00:42 AM

LoL, run along little troll, you arent welcome here.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
October 26, 2001 8:05:15 AM

... I try and bring a little bit of humour and fun to this otherwise dreary board... *sniff*
Anonymous
a b à CPUs
a b V Motherboard
October 26, 2001 9:13:55 AM

Dont worry Matisaro is a troll too.

-Spuddy

<font color=red>Being Evil Is Good. Cause I Can Be A Prick And Get Away With It.</font color=red> :lol: 
October 26, 2001 9:15:37 AM

yeah I could tell
October 26, 2001 12:35:57 PM

What are you implying? I mean if Tom said that the board support the whole thermal protection lark, then the boards documentation probably suggested that it did and the mother board manufactuerer would have said that it did. He wouldn't just "say" that it did with out any substantial evidence. I inferred from your post that you're suggesting he did do his experiement properly. Maybe I am reading to deeply into something that isn't there. My point is that if the board manufacturer lied to Tom then it can't be his fault, but at the end of the day does it matter? I mean like you all said, we routinely rip off our heatsinks for kicks.

Democracy Bernad, it must be stopped!
October 26, 2001 12:39:17 PM

Cheer up,

I think there is truth in what you said, there are such things as AMD trolls as well, it works both ways.

Charlie

Democracy Bernad, it must be stopped!
October 26, 2001 12:45:54 PM

You know, I had a whole heap of flame here, but its not worth it to give you what you want, so I wont respond to your attacks spud.

As for intel inside, his own website says he is here merely to cause trouble, so I will let that stand as it is.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~<P ID="edit"><FONT SIZE=-1><EM>Edited by Matisaro on 10/26/01 06:27 AM.</EM></FONT></P>
October 26, 2001 12:48:20 PM

No, this is what im saying.


Tom did his review and claims the mobo had thermal monitoring hardware, and he also states the mobo manufacturer said so as well.

However, many other tech review sites claim the motherboard he used did NOT, and none (which I have read) have said, no wait it did toms ok. What I am saying is that the only one who is agreeing with tom is tom, and everyone else who is speaking out about it(in the hardware community again that which I have read) has been saying toms test was inaccurate and poorly setup due to that fact.

That is all, I personally have not gone back to see which mobo he used for his test although I am considering it now.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
Anonymous
a b à CPUs
a b V Motherboard
October 26, 2001 2:14:36 PM

Actucally have anyone downloaded that video in my first post ? Here is some more mirror sites if that one is too slow.

ftp.kaosoverlords.net/pub/amdburn.zip
http://www.kaosoverlords.net/amdburn.zip

have fun and turn up your speaker :) 
October 26, 2001 3:02:59 PM

he's a troll no doubt, he's a corparate troller and fud spreader. he works for Fujitsu but thinks he's an AMD employee. !OAMLFTOR

"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"
October 26, 2001 3:25:38 PM

the vid looks like propaganda from the "Underdog" regurgitated food for you puppies.

the board Tom used is the<A HREF="http://www.fujitsu-siemens.com/rl/peripherals/mainboard..." target="_new"> Fujitsu-Siemens D1289</A>

<i>"Thermal Management
A microcontroller developed by Fujitsu Siemens reliably protects your PC from overheating, thus preventing loss of data and damage to the processor. Innovative ventilation and thermal monitoring also reduces noise. If the processor should overheat even with maximum ventilation, <b>it is automatically slowed down so that the system remains stable.</b> The microcontroller functions independently of the operating system and the processor. To fully benefit from these functions, we recommend installing the SystemGuard package."</i>

Damn that sounds like Throttling, that never worked! Fujitsu what garbage are you putting out? LOL!


"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"
Anonymous
a b à CPUs
a b V Motherboard
October 26, 2001 3:27:02 PM

I have. Im sceptic though. Can you supply a link to some offical statement from AMD that they created such a movie ? The author from the movie is reported as 'Ben & Joe', while the copyright is AMD's.

I also find it hard to believe how bad the removal of the HSF was filmed.. you dont even get to see it at all, as the guy's arm is in front of it.

Either way, true or false.. this whole thing is non issue to me anyway.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
October 26, 2001 4:54:00 PM

the video <i>IS</i> doubtful, or badly recorded.

here again is some <A HREF="http://www.64bits.org/cpuheat/cpuheat1.htm" target="_new">after analysis</A>.

where do we get the answers?

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>
October 26, 2001 5:28:36 PM

<font color=blue>"I try and bring a little bit of humour and fun to this otherwise dreary board... *sniff*"</font color=blue>

You should follow Spud's lead. He was a tremendous troll at first, but now he has all but disappeared.....


<font color=blue>This is a Forum, not a playground. Treat it with Respect.</font color=blue>
October 26, 2001 6:56:33 PM

Indeed, it is prettymuch a non issue.

From my perspective, I can see where the motherboard might have a built in thermal circuitry/thermal diode. That doesn't mean the board THG used actually utilized the thermal protection on the Palomino chips. Indeed. it seems quite plausible that the board manufacturer misunderstood the THG tester's question about the thermal protection, and was simply referring to the protection the board itself offered, which apparently was not sufficient to handle a full heatsink fan removal while in operation.

So, the whole thing is moot IMO. The Pentims were tested on boards that support the Pentium thermal protection; the Athlon likely was not tested on a board that supports that Athlon thermal protection. The incorrect information could easily be attributable to misunderstanding between the THG tester and the board manufacturer.
Mark-

When all else fails, throw your computer out the window!!!
October 26, 2001 6:56:35 PM

Indeed, it is prettymuch a non issue.

From my perspective, I can see where the motherboard might have a built in thermal circuitry/thermal diode. That doesn't mean the board THG used actually utilized the thermal protection on the Palomino chips. Indeed. it seems quite plausible that the board manufacturer misunderstood the THG tester's question about the thermal protection, and was simply referring to the protection the board itself offered, which apparently was not sufficient to handle a full heatsink fan removal while in operation.

So, the whole thing is moot IMO. The Pentims were tested on boards that support the Pentium thermal protection; the Athlon likely was not tested on a board that supports that Athlon thermal protection. The incorrect information could easily be attributable to misunderstanding between the THG tester and the board manufacturer.
Mark-

When all else fails, throw your computer out the window!!!
October 26, 2001 7:57:53 PM

Note to all:

AMD's thermal diode coding is controled by the Northbridge and Southbridge (NB/SB) on the Motherboard. If the CPU starts to overheat the diode sends the info to the NB/SB. This is a fine and dandy but if the NB/SB is not quick enough it can lock up the CPU and Fire-City. Intel has the instructions on the CPU-die. So if it needs to ramp down the power it can. AMD has to go through the chipset to NB/SB. If the main bus (voltage) is left open while waiting for the shut-down signal the CPU is still suseptible to fry.

Hopefully they will fix the instruction set into their next die/CPU to prevent problems like this. I personally like Athlons and Pentiums. I like each for it's own reasons. This is one area that Intel was smart for addressing.

<b>Your mother puts license plates in your underwear. How do you sit? - Real Genius</b> :lol: 
October 26, 2001 7:57:55 PM

Note to all:

AMD's thermal diode coding is controled by the Northbridge and Southbridge (NB/SB) on the Motherboard. If the CPU starts to overheat the diode sends the info to the NB/SB. This is a fine and dandy but if the NB/SB is not quick enough it can lock up the CPU and Fire-City. Intel has the instructions on the CPU-die. So if it needs to ramp down the power it can. AMD has to go through the chipset to NB/SB. If the main bus (voltage) is left open while waiting for the shut-down signal the CPU is still suseptible to fry.

Hopefully they will fix the instruction set into their next die/CPU to prevent problems like this. I personally like Athlons and Pentiums. I like each for it's own reasons. This is one area that Intel was smart for addressing.

<b>Your mother puts license plates in your underwear. How do you sit? - Real Genius</b> :lol: 
October 26, 2001 8:41:40 PM

Well, regardless of the coding of the diode, the question remains, "Did the motherboard that THG used in the article actually support the Palomino didoe, or was it using, as most other motherboards do right now, a thermal monitoring chip that is not directly connected to the CPU, and therefore not always the best measure of the CPU's actual tempeture?" Now, my information about the thermal diode on the XP is that it's simply a way for the motherboards to read the core temp. I can't remember anywhere in my readings it being stated that it was an actual protection measure that would shut off the CPU in case of HSF failure, or other possible problems. So, that still leaves the working of the protection to be dealt with by the motherboard chipset, which the KT266A did not support, so it would be safe to say that the KT266 (a beta version of the KT266A really) did not have the ability to read from the diode either. This is all heresay on my part, with no substance or links to back it up besides the Anandtech article. But, if history repeats itself, it will be a long time coming before the diode is supported by the chipset.

Until then, <b><i>"DON'T TAKE OFF YOU HSF FOR ANY REASON WHEN THE SYSTEM IS RUNNING!!!"</i></b>

Ahem, this has been a public service announcement by:

-SammyBoy
Anonymous
a b à CPUs
a b V Motherboard
October 26, 2001 10:18:27 PM

Oh come on Matisaro respond to me show them what a troll you are i know you want to i know you need to its your nature and AMDMeltdown is right dude your not even a real AMD employee your a Fujitsu lackey *spud giggles*. Come on man we need to some action back in the forum you and me bubble head.

-Spuddy

<font color=red>Being Evil Is Good. Cause I Can Be A Prick And Get Away With It.</font color=red> :lol: 
October 26, 2001 10:38:07 PM

Didn't Fredi ask you to stop posting? :) 

<font color=orange>Quarter <font color=blue>Pounder</font color=blue> <font color=orange>Inside</font color=orange>
October 26, 2001 11:51:17 PM

Good Article girish.......
I know you musta wrote it considering it says to mail you at GIRISH@64bits.org or whatever...

Anyways, very valid issues there, and Let's see if Mr. Lazzy ass Tom will ever respond to it.....

-MeTaL RoCkEr
My <font color=red> Z28 </font color=red> can take your <font color=blue> P4 </font color=blue> off the line!
Anonymous
a b à CPUs
a b V Motherboard
October 27, 2001 12:03:03 AM

Ya but im bored and looking for a new job so why not stir up some troubles.

-Spuddy

<font color=red>Being Evil Is Good. Cause I Can Be A Prick And Get Away With It.</font color=red> :lol: 
October 27, 2001 4:34:25 AM

Apparently spud and meltdown are trying to goad me into flaming them, but I will not take part in their childish game.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
October 27, 2001 5:13:38 AM

Guess it's safe to say that Palomino does not support any kind of thermal protection. To rephrase that, the Palomino only supports internal core thermomonitoring. It is still the duty of a external controller to shutdown the CPU. It's kind of like programing your system's BIOS to shutdown the computer when the temperature gets too high except that the mainboards chipset reads off of the internal diode inside the CPU's core which results in a more accurate and faster response time over the current mainboard thermosensors.

<font color=blue>Another waste of bandwidth on the web. :tongue: </font color=blue>
October 27, 2001 7:16:32 AM

Quote:
AMD's thermal diode coding is controled by the Northbridge and Southbridge (NB/SB) on the Motherboard. If the CPU starts to overheat the diode sends the info to the NB/SB.


Actually, the Athlon manuals do not say this. The diode leads are brought out so that a thermal sensor/control on the motherboard could monitor the temperature. They arent connected to either NB or SB in anyway.

A number of manufacturers make NBs and SBs for AMD processors, and there is no standard, as far as I know that specifies how the Bridges' react to temperature, the Southbridge is just an interface to the PC HW monitor ASIC chips, which collect data from external sensors. But afteall its totally an external circuit.

See <A HREF="http://www.64bits.org/cpuheat/cpuheat1.htm" target="_new">this article</A> for datasheet references and detailed analysis of the test.

The Palomino/Morgan or later processors have this thermal diode which should be connected to a thermal control circuit on the motherboard, which could be external to any of the bridges or even the LPC/HW monitor that connects to the NB/SB. This Thermal Control circuit should monitor the temperature and manipulate the FSB or multiplier (which can be easily controlled by the BIOS), but I think it should be a separate chip that can make decision by itself even when the BIOS is not functioning. It could make a better Athlon board! Or if such circuit is integrated into the Athlon die itself it would become as indestructible as P4! It could be programmed by the BIOS or Windows or will have its oqn defaults.

One more thing - a rise of temperature to as high as 370°C in 5 seconds is staggering. It amount to almost <b>70°C/sec</b>!!! Is it practical? Or THG guys have taken us all for a ride? See <A HREF="http://www.64bits.org/cpuheat/cpuheat3.htm" target="_new">detailed analysis of this stunt</A>! Any answers to the questions asked?

girish

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>
October 27, 2001 2:37:47 PM

Quote:
Guess it's safe to say that Palomino does not support any kind of thermal protection. To rephrase that, the Palomino only supports internal core thermomonitoring. It is still the duty of a external controller to shutdown the CPU. It's kind of like programing your system's BIOS to shutdown the computer when the temperature gets too high except that the mainboards chipset reads off of the internal diode inside the CPU's core which results in a more accurate and faster response time over the current mainboard thermosensors.


Exactly.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
October 27, 2001 3:09:24 PM

thats my spuddy

Nice Nvidia and ATi users get a Cookie.... :smile: Yummy :smile:
October 27, 2001 3:18:24 PM

<i>Guess it's safe to say that Palomino does not support any kind of thermal protection. To rephrase that, the Palomino only supports internal core thermomonitoring. It is still the duty of a external controller to shutdown the CPU. It's kind of like programing your system's BIOS to shutdown the computer when the temperature gets too high except that the mainboards chipset reads off of the internal diode inside the CPU's core which results in a more accurate and faster response time over the current mainboard thermosensors.</i>

to sum it up: the AthlonXP/MP has half-assed thermal protection.


"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"
October 27, 2001 3:27:33 PM

AMDmeltdown i have a question. Say if you can get a FREE Athlon XP 1800+ or spend $500 for a P4 2.0ghz. What would you take ????

Nice Nvidia and ATi users get a Cookie.... :smile: Yummy :smile:
October 27, 2001 3:49:02 PM

I guess it is better to have a dedicated chip that would decide by itself whether to reduce the FSB/Multiplier or shutdown the system.

There is no point in relying on BIOS since in most overheating events, the processor simply freezes. Then, there is no BIOS running to shutdown the machine!

girish

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>
October 27, 2001 4:07:31 PM

>AMDmeltdown i have a question. Say if you can get a FREE Athlon XP 1800+ or spend $500 for a P4 2.0ghz. What would you take ????

first of all, I will never camp out like a AMDpuppy for a pos cpu from *md, however! :)  if I turned the corner and I saw chimpy handing out stuff, then of course I'd take it. then, I'll build a system and price it above a P4 sell it to get maxium return from the ignorant pups.


"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"
October 27, 2001 4:15:09 PM

hmmmm... WTF with the monkey and puppy's

Nice Nvidia and ATi users get a Cookie.... :smile: Yummy :smile:
October 27, 2001 4:33:33 PM

AMDMeltdown, you realize that the Athlon XP 1800+ is faster than the P4 2GHz in most tasks right? I don't care about cooling 'cause I'm 15 and I got my Athlon system to work the first time at good temps, so don't say the Athlon is fragile because it's not. Stability isn't an issue either. My Athlon is rock solid with a mild overclock under Windows 2000. I can sometimes leave it running for days without problems. AMDMeltdown, you're a hypocrite. You claim AMD fans are ignorant when, in fact, it takes more research to find out Athlon XPs give you more bang for your buck than Pentium 4s, because AMD processors aren't yet as popular as Intel processors.



AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor
October 27, 2001 4:37:17 PM

AMD man, you should just ignore him and everyone else like him, because the general forum community takes what they say with a grain of salt, I used to get pissed when they attacked me or spewed their fud, but it does nothing but make you look worse to sink to their level by catering to their trolldom with replies. Dont feed the fire, let the trolls have their party, it dosent hurt us or the forum.



~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~
!