PDA

View Full Version : Is my PC too hot?


DaForce
03-19-2003, 04:10 AM
Hiya ppls,

I was just wondering if my pc was too hot, the reason i say this is originally when i got it, it was running about 38-41*C unloaded, and about 43*C loaded, and now it is about 43*C normal use...websurfing....dreamweaver work...stuff like that. and up to 47-48*C underload..heavy 3dsmax renders. (temps reported via Asus Probe)

The only thing that i think could have cause it is i put 2 more harddrives into the removable harddrive bay, on my Lian-Li PC60 case, now this bay, sits right infront of the front two case fans (it sits on the bottom of the case), so im thinking that i have restricted the airflow quite abit....also these hard drives are not plugged in the just sit there until i need them. so they are not heating up at all.

ahh yeah...my system specs.
AMD 2700
SLK800 with thermaltake smart fan2, set to thermal control, it sits at around 3300rpm....(max speed is 4800)
artic silver 3
Lian-Li PC60 Alluminum Case
Antec True Power 550w PSU
ASUS A7N8X mobo
2x512 Corsair XMS3200
Powercolor Radeon 9700pro
SB Audigy2
Western Digital 80gb w/ 8mb cache
plus the 2 other drives...which are basically just taking up space at the moment.

Ahh yeah also the case temp is almost always with 5-7*C of the cpu temp...and my room is usually pretty warm...i guess about 23*C ish.

I find it odd that the cpu temp would run so much, but if it is because of the other drives blocking the airflow them i shall move them.

p.s. the comp is only a month or so old.

p.s.s. i realise my computer is in no way hot enough to damaged anything, but its odd that the temp jumped.

Thanks y'all

gabe28
03-19-2003, 05:04 AM
Those are perfectly good temps to be running at for an AMD. In fact you only really have to start getting concerned if your temps get above 60C. To be on the safe side you can probably set your computer to shut off if it reaches a certain temp. This is done in bios. Judging by your number though, I don't think you have anything to worry about. My own system idles at 48 to 51C but I only have one tower fan right now. It's completely stable though. I'll put some more fans in to be safe.

DaForce
03-19-2003, 05:10 AM
ive got plenty of fans...6 in total, 5 of which are 80mm. But i think the reason the temp jumped is because i blocked some of the air flow with those harddrives, plus i need to neaten up the inside of the case, cause there are cables going everywhich way.

I will try moving the harddrives and see what the outcome is.

MaDSheeP
03-19-2003, 05:53 AM
dust can drastically effect the performance of your heat sinc...

buy a can of compressed air, and clear things out in there... my idle temp dropped 5 degrees C when i did that =)

DaForce
03-19-2003, 06:01 AM
wow, 5*C hey, i will look into that, as my room sometimes gets just a tad dusty.....but hey dont they all.

but it is a fairly new machine, and last time i looked, a few days ago, it was all nice and clean. But you never know.

Thanks

Valkyrien
03-19-2003, 06:33 AM
I've never seen my Athlon 2400 run above 43C...then again it's never gotten past BIOS period, so :shrug:

DaForce
03-19-2003, 06:40 AM
never gotten past bios!!! what the.....

why is that?

GregHess
03-19-2003, 11:40 AM
Daforce,

Those are fantastic temps. If you've got a adjustable rpm fan array, you might even be able to tweak rpm's a bit. (To reduce noise if the system is loud)

I try to tweak systems so that their load temps fall around the 50C mark. (Which is exactly where your system is hitting underload) Usually results in pretty quiet systems, as long as you use oversized/high performance heatsinks like the thermalright you got.

Valk,

Do you want to try starting another thread on your athlon system? So we can try and help ya again?

DaForce
03-19-2003, 11:50 AM
GregHess,

Yeah my machine is very quite, very quite indeed. basically cant here it once i have any sort of background noise...e.i. Tv on quite, or a fan of some sort.

Im mainly curious as to why the temps jumped up by 5*C, but i reckon it has something to do with the restricted air flow since i put in those 2 harddrives

GregHess
03-19-2003, 12:25 PM
Well if you placed 2 HD's in there, its going to present a two fold problem.

1) The hd's blow the incoming airflow some.
2) Each HD is a heat source for the machine. Depending on the type and speed, drives can get upwards of 100F (er 38C?). Stick two in there, and you get the idea.
3) HD's require juice to run, the more juice the system needs, the harder the psu works, and thus the hotter.

Sounds like the system is still fine though, well in the green. Actually its pretty far away from yellow :).

DaForce
03-19-2003, 12:33 PM
actually the harddrives are not even plugged in at the moment, cause i only plug them in when i need them , so actually its just a single fold problem :-) I will move them and see if that helps lower the temps.

Either way like you said it is still well in the green.

GregHess
03-19-2003, 12:34 PM
Df,

Have you thought about getting external converters for the drives? Would be a bit more convient then having to constantly plug and unplug them in (having to open that case that is). Newegg sells some external converters which work with most IDE drivers. They come in both firewire, USB1.1, and USB2.0 (which would work great with the nforce2 board).

DaForce
03-19-2003, 12:39 PM
Im actually going to be buying the Lian-Li Removal harddrive bay, it looks great is alluminum and supports up to 10,000rpm drives. So it will look pretty cool.

I would rather not have external devices, but i may have a look into them just because of the ease of use factor.

GregHess
03-19-2003, 12:43 PM
Ah I had forgotten about the removable drives. That would be a great solution as well. I've used the RH3200 (I think) series for 7200 rpm drives. Actually used them with a mirror raid array to create a forced raid failure (thus allowing the user to clone an entire drive once a month without too many steps).

The nice thing about the drives is (in your situation), if you turn the key to unlock, the drives wont' recieve power, even if their in the bay. So in that manner you could control whether the drive was in use or not.

Do remember that their not hotswappable however, and it'll take a full system shutdown to turn the drives on.

DaForce
03-19-2003, 12:54 PM
raid is always a good thing.

Yeah being able to control what has power and what doesnt will really help keep things cool.

And yes i know that they are not hotswappable....that stuff is common knowledge....or atleast it should be. I have been using removable drives for about 3-4 years now. Mainly for when i was testing 98,NT, ME and 2000, and i didnt want to risk have flaky dual boot drives. So i just got myself a few drives and away i went. But the removable drive bay i got back then was pretty cheap, only about $25AUS, but it did the trick. Whereas the Lian-Li one im getting will cost about $170AUS i think, and additonal HDD carriages are $80AUS...i think.

Ah well it will be worth it.

DaForce
03-19-2003, 01:01 PM
Actually Greg, if i can just ask you a quick question.....about swap files under Win2k. I have setup a partition on my HDD purely for windows swap file, i have apparently set it up under ...control panel....system....advanced tab....performance options....virtual memory....and everything seems ok...but when i was doing a "hefty" render in max, which stated 400 or so of physical and 700 or so of virtual memory was in use, the used space on said swap drive was at the minimum swap file size i had set (128mb) i check under task manager and there was aparently quite abit of swap being used, but this swap file was not growing any larger in size....do i have to change something somewhere else to make the change work???

I will attach an image of my swap file settings

http://www.apex.net.au/~max3ds/images/swapfile.jpg

Thanks Heaps

GregHess
03-19-2003, 01:25 PM
Hey DF,

Its usually recommended to set your swap file min and max to the same number...usually the same as your physical ram allotment. This prevents your swap file from growing/shrinking, thus reducing the amount of fragmentation that's occuring. Ex with 512 megs of ram, min 512, max 512. Give it a try and see if it helps/hurts.

Did you check the C:\ drive to see if its size was increasing? I know in some instances, even if you specify a different drive/partion as the swap file, it still does the swapping on the system disk.

DaForce
03-19-2003, 01:29 PM
so basically i did everything that i should have done to change the swap file location, right?? But its still possible that its still using the system drive as the swap drive...is there anyway of overcoming this?

I have not checked to see if the c drive changes...is there n actual file for the swap file...like in windows98 it was win.swp or something....is there a similar thing in win2k or is it completely hidden?

Ta Greg

GregHess
03-19-2003, 01:37 PM
I'm not entirely sure. I just know I've seen alot of threads on the subject, and it actually seems impossible to utterly and completely force windows to move the swap. Even if you have a seperate drive and EVERYTHING points to its swap, windows will still have a tiny swap partion on whatever partion the system files are on. Why? I have no freaking clue, blame microsoft.

On a side note, it doesn't seem to make much performance difference moving and optimizing the swap file all over the place. Whatever works, use :).

I would set the min max to whatever your physical ram allotment is, and just watch the system disk for changes in size while rendering, to make sure its utilizing the swap on your secondary partion.

Note: You can also setup performance counters (under admin tools) to watch the swap file's usage and the c:\ d:\ access.

DaForce
03-19-2003, 01:44 PM
bloody windows :-)

Ok will do, i will keep an eye out next time i get a heavy render going.

BTW i am just doing a defrag on my C drive and it was quite fragmented, here is what is says about my scratch disk (E drive).

Volume fragmentation
Total fragmentation = 17 %
File fragmentation = 35 %
Free space fragmentation = 0 %

Is this abnormally high??......i have had the computer about a month. And have not yet defraged any of the partitions...but i am doing them all now.

GregHess
03-19-2003, 02:20 PM
Did you defrag after you setup the machine?

By default a new install is horribly fragmented. Should always defrag after you finish updating, installing all apps, updating drivers, and cleaning out all the temp files.

Valkyrien
03-19-2003, 08:11 PM
Originally posted by GregHess

Valk,

Do you want to try starting another thread on your athlon system? So we can try and help ya again?

Not really...I kinda gave up and took it to a local computer repair place...could be awhile though, the guy there seemed to be about as mystified as I am!:shrug: needless to say though, I'll let everyone know as soon as it works again!:)

gmask
03-19-2003, 08:54 PM
According to the white paper from AMD
http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/25175.pdf

The 2700 can run to a max of 84c so you are well below that. Older models can run up to 95c.

GregHess
03-19-2003, 09:04 PM
gmask,

Just a quick note....

The recommended temps mentioned in this forum are for maximum stability. Once you start getting around 60C there is a possibility of errors occuring, which can/will cause system instability and possibly a crash.

The hotter it gets, the greater the likelyhood.

Although the chip might not get damaged for another 20C, that doesn't help much when you miss the clients deadline because of a system crash caused by an overheating processor.

gmask
03-19-2003, 09:32 PM
Originally posted by GregHess
gmask,

Just a quick note....

The recommended temps mentioned in this forum are for maximum stability. Once you start getting around 60C there is a possibility of errors occuring, which can/will cause system instability and possibly a crash.

The hotter it gets, the greater the likelyhood.

Although the chip might not get damaged for another 20C, that doesn't help much when you miss the clients deadline because of a system crash caused by an overheating processor.

I'm not saying that you want it to run at max but unless you are over clocking then I don't see the concern. From my experience running many very hot computers during the summer they don't crash and if they do it has more to do with user error than CPU error.

DaForce
03-19-2003, 09:33 PM
GregHess: ahhh no, i "neglected" :-) to defrag my machine after the initial install, that would probably explain the fragmentation. The thing i noticed is that movie files seem to get really fragmented. For instance, i download a whole bunch of Battlefield 1942 stunt movies, and these were all on my scratch disk, now these files were the ones that were most fragmented, some over 200 fragments. Is that because of the nature of movies and the way they play off the disk??

gmask: 84c, man i cant imagine that would be very stable. It probably says in fine writing...These chips "can" run at 84c but only for a period of 30sec, at which point they implode, taking the heatsink and motherboard with them.

GregHess
03-19-2003, 09:42 PM
Originally posted by gmask
I'm not saying that you want it to run at max but unless you are over clocking then I don't see the concern. From my experience running many very hot computers during the summer they don't crash and if they do it has more to do with user error than CPU error.

I'd say...

At 60C a system is 25% more likely to crash then at 55C. At 65C its 50% more likely to crash then at 55C. Most default motherboard cpu alarms will go off at 70C. I believe the default shutoff temp is 75C.

Hey its your paycheck. I for one would rather have insurance.

gmask
03-19-2003, 09:47 PM
Originally posted by DaForce
gmask: 84c, man i cant imagine that would be very stable. It probably says in fine writing...These chips "can" run at 84c but only for a period of 30sec, at which point they implode, taking the heatsink and motherboard with them.

You're at half of the MAX I doubt that you would have any errors at 10 or 15 degrees above that. Your computer should be set to shut off if it gets to the max temp or within 5 degrees of it. Most modern processors will fry if it runs for 10 seconds without cooling...as in no fan and no heatsink. So even if your fan dies the heatsink and the limit set in your bios should prevent damage from occuring.. BTW if this were to happen at the worst you might hear a popping sound.. no flames or smoke and definatly no explosions.

DaForce
03-19-2003, 09:49 PM
gmask: Yeah user error and incompetance, are the 2 most deadly forces in the computer world. How many times have you seen someone buy a computer with fairly low end components, and then 3 months down the track, after they have installed every little app on every pc mag cd they can get their hands on, and littered their harddrive with misplaced files, they bring their computer back and say "Why is my computer running so slow, i have not used it that much"!!! These are the sort of people who dont understand things like:
"Un-Install"...they just go and delete the program folder because this is a much easier and efficient way of removing programs...
"File Management"...Well there is no need to install things in their own folders, because i like the ease of browsing thru 1 big directory of 10,000 files.
"Common Sense"....Oh look here is an email from somebody that i dont know....oh look there is an attachment...i had better open it just to make sure its not important...then they come in and say stuff like " i think someone hacked my computer and put a virus on there"..needless to say i have to restrain myself from wanting to beat them upside the head with a trout.

Anyway, back on track now, Yeah like Greg said, they hotter things get the more unstable things will be, and of course think about if a cpu is running at 75*C for instance, then how much do you reckon that would effect the surrounding hardware, now days, video cards are fairly close to the cpu...and they tend to not like a lot of heat. Basically running at higher temps, strains the whole pc thus shortening its life.

DaForce
03-19-2003, 09:54 PM
gmask: I realise that there would be no explosions or "implosions" i was just be a little humerous :-)

gmask
03-19-2003, 10:27 PM
Originally posted by DaForce
Anyway, back on track now, Yeah like Greg said, they hotter things get the more unstable things will be, and of course think about if a cpu is running at 75*C for instance, then how much do you reckon that would effect the surrounding hardware, now days, video cards are fairly close to the cpu...and they tend to not like a lot of heat. Basically running at higher temps, strains the whole pc thus shortening its life.

I dunno.. I have hardly seen a computer die affter being abused to the hilt with heat, cigarrette smoke and malfunctioning users.

I will grant that I think my video card is the source of most of my crashes when I have them. Then also your chances of crashign increase as you add more components to the system but not because of heat but rather due to drivers and incompatibilities. Processors become obsolete in about 6 months after you buy them so I so I don't really fear heat wearing out my machine in the next 5 years.. The oldest computers I have have passive heatsinks and they seem no where near being dead but they are also nowhere near as fast as a modern processor. But if you think that you computer will feel as fast as it does now two years from now I'd be surprised but I have also known people who thought their computer was actually working slower because it had been worn down over the years.. I dunno I have not read any studies on this.. I guess it is possible.

Anyway.. I'm not a CPU designer or anything but I think if you are under the MAX by ten degrees that you are not going to see a significant increase in errors. It sounds a bit paranoid to me to worry about a 5 degree increase in temp. I would look first to other sources of problems. Usually when people say that they are having problems I ask them about the computer specs but not what temp are your CPU's running at.

Since AMD's run normally at a higher temp than Pentium's I think there has been this irrational concern about the AMD's running at their normal temps.

Then on the super paranoid side of things there are those who will swear that on a stable system that a crash is actually a subatomic particle passing through the CPU and causing it to fail.

Actually the only time I have seen a computer smoke was a power supply that got a static shock.. it fried and gave out a nice puff of smoke. The irony being that you are typically told to discharge yourself by touching the power supply when installing ram or CPU's.

Oh and then there was the time lighting fried my synthesizer keyboard and it smoked too. Then there was the time I stuck a fork in the toaster..just kidding.

GregHess
03-19-2003, 10:44 PM
Here is some more comments.

Processor cooling is becoming one of the most important upcoming decisions in computer manufacturer. Millions of dollars are current being expended to try and find a better way of cooling central processing units. At the current rate of processor wattage dispersion, we'll be looking at 200 watts/cm in just a few years...if not sooner. For those who aren't aware, the surface temperature rocket exhaust is about 1000 watts/cm.

Current technology will not support air cooling much longer, and an alternative mean must be found to maintain a cool running chip.

Here's a short blurb on the subject.

http://www.overclockers.com/tips00315/

With this information at hand...

Why would so many manufacturers spend so much time and money trying to lower the temperature of cpu's? Stability and Speed.

The cooler you can maintain a chip, the faster it will go...its even been proven that a CPU running at extremely lower speeds runs FASTER at the same clock then the EXACT SAME cpu at a higher temperature.

Here's an article on that subject, as well as a review of a unit which performs it.

http://www.tech-report.com/reviews/2003q1/asetek-vapochill/index.x?pg=1

(Though his results were similar at both temps, the theory is still useful to read for this discussion)

So a bunch of these people think its realitively important to make sure the cpu is running cool....But why is it important?

http://www.overclockers.com/tips30/

Keeping a processor cool....

1) Increases system stability.
2) Prolongs processor life and durability.
3) Allows for more give/take when harsh conditions arrive (summer).

Here's another article on the subject...and a massive cooler review.

http://www.dansdata.com/coolercomp.htm

You'll also find that both intel and amd have white papers (more exhaustive then the ones you mentioned) detailing the heat problems facing processors today.

Intel's processors are actually as vulnerable if not MORE vulnerable to heat damage then AMD's. Their latest processor (the 3.06 HT P4) actually has a throttling temp of 69C, which is the temp at which the processor begins reducing its gigahertz...to prevent damage.

So in conclusion...

Why should you look into additional cpu cooling if your temp is hovering around 60C or greater?

1) System instabilty increases past 60C on both AMD and Intel boxes. This is due to heat acting as resistance to the electrical circuitry in the core. Too much heat and errors can occur...some errors and your l2 ecc should catch them...too many and the system crashes.

2) Processor longevity. The hotter the chip is running, the greater the reduction in the processors lifespan. Though if you upgrade frequently...once every year or every other year, this might not be a big deal...it becomes a big deal in a 3-5 yr span of time, especially if the unit is on 24/7 under stress.

You might of had some luck with your cpu's running under extremely adverse conditions...but my experience has been (over 5 years now) that if the chip is in the 60C's, you better not be relying on it.

Thats my 2 cents and I'm stickin too it.

DaForce
03-19-2003, 10:50 PM
gmask: haha yeah that is very ironc, same thing happened to me one day, plugged in the PSU and there was a funny buzzing noise....but it was not that strange, 5 secs later, a very loud POP, and flash of blue light, and a little puff of smoke and the PSU was dead. hehehe fork hehehe toaster....i bet the same people i mention before with their lack of computer knowledge would be the sort of people to actually stick a fork into a powered toaster.

Greg: Wow, i think you have definitly closed the book on this conversation. i definitly agree, higher temps = lower performance = system instability = shortened life of parts.

gmask
03-19-2003, 10:58 PM
>>>So a bunch of these people think its realitively important to make sure the cpu is running cool....But why is it important?

http://www.overclockers.com/tips30/


Because they are running their procs above spec..so yeah it would be really important to reduce heat.



>>You'll also find that both intel and amd have white papers (more exhaustive then the ones you mentioned) detailing the heat problems facing processors today.

Could you give a link to the one for AMD.. I'd like to read it espcially if it is better than the link I gave.

>>>Intel's processors are actually as vulnerable if not MORE vulnerable to heat damage then AMD's. Their latest processor (the 3.06 HT P4) actually has a throttling temp of 69C, which is the temp at which the processor begins reducing its gigahertz...to prevent damage.

Didn't I basically say that Pentium's required a lower normal operating temperature? The AMD's due to their design run hotter than a comparitable Intel proc.



>>>You might of had some luck with your cpu's running under extremely adverse conditions...but my experience has been (over 5 years now) that if the chip is in the 60C's, you better not be relying on it.

If lives were in the balance or millions of dollars then I'd get a cold room. Back to my main point which was that if the max is 85 and you are running at 40 then you can safely say that if you are running %50 or more below the max spec then you ahve little to worry about. jeez. But if you want to get a water cooler for your computer, freon injectors or flux capacitors be my guest.


>>>Thats my 2 cents and I'm stickin too it.

Great thanks for the info.. there is no need to get hot under the collar..pun intended ;-)

gmask
03-19-2003, 11:01 PM
I think I'm going to install hair dryers in all my computers and turn on a heater even though it's like 70 degrees out and have the slowest hottest computers around. Then I'll call it a sauna for computer nerds and make millions of dollars to the sound of popping CPU's. :bounce:

DaForce
03-19-2003, 11:29 PM
gmask: hahahah you could make a fortune :-)

gmask
03-20-2003, 01:21 AM
Originally posted by DaForce
gmask: hahahah you could make a fortune :-)

I went to this party once and they had an area off to one side with a pile of old monitors and a couple of basebell bats, eye protection and gloves and you could beat a couple of monitors senseless for fun. Nothing like breaking stuff to releave stress ;-)

Valkyrien
03-20-2003, 01:24 AM
hmmm, speaking of which, there's an old broken monitor in the basement...and it can't be hard to scare up a baseball bat....*evil grin*

DaForce
03-20-2003, 01:38 AM
hahahaha you cant beat sensless computer violence.

A guy i knew once, he was a bit of a psycho, couldnt get his computer working properly, so he took out the harddrive and smashed it on the ground about 10 times,and kicked it around for awhile, then he took out the ram chips and started smashing them with a hammer, he then tried to break the motherboard over his knee but after 5 mins of trying he gave up, his hands were all cut up from the back of the motherboard, but it was amazing he had it almost folded in on itself and it would not snap. Mind you it wouldnt be much good for anything other than a paper weight tho. The other thing was it wasnt really his computer, it was given to him to use as he did not have a lot of money, and he just could not take the fact that he would have to wait until the next day to get someone to bring in their windows cd so he could get his machine working....so he destroyed it......needless to say he was pretty "disturbed". Also this whole display of computer rage, was done just out side of a computer games arcade there was a fairly large audience, plus the arcade was situated with a large bus stop/depot kinda place.....funny stuff.

gmask
03-20-2003, 02:15 AM
Woah!

Reminds me of that video that was circulating on the net were you see this cubicle worker go off and shoves his computer off the desk. Dilbert goes psycho!

DaForce
03-20-2003, 02:20 AM
heheh yeah, except this was really. That one was fake you could even see that the keyboard and mouse were not even plugged in. But still pretty full on stuff.

DaForce
03-20-2003, 04:07 AM
On another note, i just got a call from my Mum, and she said that our house had just been broken into. But the good news is that my 3 computers were not stolen...phew...but my gamecube and xbox were nicked :-(

'bout to go home and check out the damage.

gmask
03-20-2003, 04:25 AM
Originally posted by DaForce
On another note, i just got a call from my Mum, and she said that our house had just been broken into. But the good news is that my 3 computers were not stolen...phew...but my gamecube and xbox were nicked :-(

'bout to go home and check out the damage.

Wowsers!.. My girlfriend was hit by a car several hours ago .. she's okay though..might have a cracked rib! Uh weird/bad day to say the least!

jbradford
03-20-2003, 04:42 AM
:lightbulb

big thing i just discovered on cpu temp is dust collecting between the heatsink and fan. i was hitting 63c and am now at 54c after i cleaned it. best way to do it is with a toothbrush because it wont push down dust down into the heatsink fins. my guess in drastic temp change is because the dust insulated the heat from the sink and blocked air from the cpufan. the dust was so think that compressed air didnt work.

CgFX
03-20-2003, 09:32 AM
Originally posted by GregHess
I'd say...

At 60C a system is 25% more likely to crash then at 55C. At 65C its 50% more likely to crash then at 55C. Most default motherboard cpu alarms will go off at 70C. I believe the default shutoff temp is 75C.

Hey its your paycheck. I for one would rather have insurance.

AMD's own technical documentation says that the max die temp of a series 6 and series 8 XP is 90C.

Strange...

gmask
03-20-2003, 05:32 PM
Originally posted by CgFX
AMD's own technical documentation says that the max die temp of a series 6 and series 8 XP is 90C.

Strange...

Oh ma god you read the tech docs on the AMD site .. for shame.. you know you need to get your computer liquid cooled or it might implode!!!

In 16 years your computer might not be as fast as it is today if it runs too hot and you might need to buy another one..oh dear!

GregHess
03-20-2003, 08:22 PM
I thought this conversation was over. Seems like it won't die. So here we go again.

AMD's own technical documentation says that the max die temp of a series 6 and series 8 XP is 90C.

I don't deny this is AMD's specification. However if you look at Asus's C.O.P. (Cpu overheating protection) system, you'll note that the default CPU shutoff temp is 75C.

http://www.au-ja.org/eng/reviews-asusa7v333-cop-e.phtml

There is also a video available showing this in action.

Many other motherboard manufacturers use a similar system, even for Pentium IV based CPU's.

They all have a similar threshold. Around 70C for the alarm, 75C for the shutoff. A few have the alarm set a bit lower by default.

Why would the system shutoff far before AMD's temperature limit? Perhaps the motherboard manufacturers research points out that temperatures exceeding 75 Degrees Celcius causes damage not only to the cpu, but surrounding motherboard components.

Continuing the debate...

On Page 17 of

http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/23794.pdf

AMD recommends that the temperature 0.5 to 1.0 inch away from the central core of the processor, be LESS than 40C. The Ambient air temp around a core running at 65C could be higher then this #.

Using ASK AMD about processor heat returns these results....

Improper Case Airflow (http://139.95.253.213/SRVS/CGI-BIN/WEBCGI.EXE/,/?St=54,E=0000000000074221694,K=1903,Sxi=7,Case=obj(1483))

Insufficent Heatsink (http://139.95.253.213/SRVS/CGI-BIN/WEBCGI.EXE/,/?St=54,E=0000000000074221694,K=1903,Sxi=7,Case=obj(1497))

Why is my system unstable? (http://139.95.253.213/SRVS/CGI-BIN/WEBCGI.EXE/,/?St=62,E=0000000000074222786,K=1903,Sxi=7,Case=obj(342))

And the kicker...

Thermal issues are often at the root of system stability problems... (http://139.95.253.213/SRVS/CGI-BIN/WEBCGI.EXE/,/?St=2,E=0000000000074222442,K=1903,Sxi=7,Case=obj(306))

These technical support posts, especially the last one, seem to indicate that AMD feels very stronger that heat can and will cause system stability, and maintaining a low, stable, cpu temperature is a fundamental component of system stability. In almost all troubleshooting issues involving instability on AMD's site, the first topic mentioned is heat, and processor temp. I hope this tends to reinforce the importance of cpu/case cooling in this discussion.

As for this comment....

Didn't I basically say that Pentium's required a lower normal operating temperature? The AMD's due to their design run hotter than a comparitable Intel proc.

This is a true statement EXCEPT for Intel's newest cpu's. The 3.06 and upcoming 3.2 HT break the normal mold of Intel's flagship line running cooler then AMD's. These chips produce as much or more heat then the current Athlon XP line. Actually it looks like even the Barton's are running cooler then the last few Intel processors...

Here's the old P3 wattage and max heat data.

http://developer.intel.com/support/processors/PentiumIII/thermal.htm#5

The Pentium IV 3.06 HT produces 81.8 Watts.
The Pentium IV 2.8 produces 68.4 Watts.

I'd link to this data, but it requires that you be a reseller to access it. (You have to apply for membership)

The 3000+ Barton produces 74.3 Watts of heat.
The 2800+ Barton produces 68.3 Watts of heat.
The 2700+ and 2600+ are 68.3 as well.

http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/26237.pdf

Page 21

http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/25175.pdf

Page 27 for the 2700 and 2600+ 333 XP's.

I hope this ends this thread.

In 16 years your computer might not be as fast as it is today if it runs too hot and you might need to buy another one..oh dear!

I would also appreciate it gmask if you would stop acting in a rather imature manner by taking others statement's and turning them into jokes. Its not exactly conducive to civil discussions on this forum.

DaForce
03-20-2003, 09:31 PM
Greg,
Wow i always thought that AMD were the biggest heat producer....that doesnt seem to be the case anymore.
On another note, it was quite abit cooler last night, and my comp was running at 40c Max, but i was not really doing anything intense on it. So maybe its 2 things, the restricted airflow and the hotter temps we had the other day. Either way it is still way way way in the green zone, i was mainly interested why the temps jumped all of a sudden.

gmask,
Just to clear things up, the reason i posted this thread was to try and work out why my cpu and case temp jumped by like 5c , im not worried that my cpu is going to burst into flame (and i realise that it cant really burst into flames), and i know that my computer is far far away from any sort of dangerous temperatures.

gmask
03-21-2003, 01:52 AM
Originally posted by DaForce
gmask,
Just to clear things up, the reason i posted this thread was to try and work out why my cpu and case temp jumped by like 5c , im not worried that my cpu is going to burst into flame (and i realise that it cant really burst into flames), and i know that my computer is far far away from any sort of dangerous temperatures.

Thanks .. you can at least appreciate a joke. Greg on the other hand will have to chill out if he wants his goat back.

All I have to say is that if you are around %50 percent of the max rating that the manufacturer specifies then everything should be A okay. I admit that I did say that you could be 15 or 10 degrees below the MAX and that would probably be too hot and cause problems. On the otherhand I also still stick by the motto of how much of a difference will it make 5 years from now let alone 10 or 15 as if you look at computers 5, 10 or 15 years ago you would not still be using them. So a little bit of heat degradation is bound to happen and so what? I doubt that those who are so concerned about such things use the same computers and proccessors for extended periods (5+ years) of time anyway. Let 'em burn ;-)

DaForce
03-21-2003, 03:23 AM
Gmask,

I just didnt want anything thinking i was a clueless paranoid whiner. Cause im not...honest :-)
And i do love a good joke...so bring em on ...:bounce:

Yeah i understand your views on the whole heat thing, i just reckon the cooler the better, since computer parts generally perform better when they are cooler. Just personal preference. s'all

gmask
03-21-2003, 03:41 AM
Originally posted by DaForce
Yeah i understand your views on the whole heat thing, i just reckon the cooler the better, since computer parts generally perform better when they are cooler. Just personal preference. s'all

Yeah I just installed PCI slot fans in my com-pu-ters and am in the process of upgrading bioses etc..spring cleaning.. cleaningout the dust bunnies and washing the filters for the front chassis fans(yuck!). I just like to nail down good rules of thumb for these things as it should not require a degree to divide a max value by 2 and come up with a reasonable frame of reference for CPU temperature.

Actually come to think of it I installed these two really hot scsi drives and while they were not effecting the overall system they themselves were not being cooled well enough by the build in fans. The solutions was to get a tiny fan that would fit in between their mounts and the case as there was not room for it to go anywhere else.. actually the fan came off a spare CPU cooler I had and works great to this day. I also have removable drive sleds for all my computers so that I can easily turn off spare drives when I don't need them.

Personally I get more concerned about my drives being too hot as if they fail then there is a real problem. Otherwise the average computer these days seem to be pretty stable to me... at least it seems that the operating systems have gotten as stable as the hardware..but I remember days when that wasn't so true.

DaForce
03-21-2003, 03:51 AM
filter washing....**shudder** **shudder**

"at least it seems that the operating systems have gotten as stable as the hardware..but I remember days when that wasn't so true"

Heheh yeah like a year ago :-) Bloody Microsloth, you would think by now they would have learnt how to produce a operating system that doesnt crash. You can "almost" blame every computer problem on windows..."why does my computer crash all the time" .."Oh its just the way windows works"

CGTalk Moderation
01-14-2006, 04:00 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.