How hot could a computer feasibly get?

Sort of thought provoking, I've been thinking about this for awhile. How hot can a computer feasibly get without any type of cooling or airflow? Assuming that nothing could melt and that you were running it under load, how hot could a laptop or even a desktop get? I know it depends on the hardware but just assume with me that it doesn't matter. Any thoughts?

 

Some people can get up to 100C with cooling, so i wouldn't be surprised with 150C+.

 

If it had no cooling or vents?

It would get hot to the point of turning off from the heat.

I recently saw a topic in the Sager and Clevo forum, where the OP's laptop had a blanket on top of it through the course of the day, and the laptop was on. When he came home and saw the blanket, he took it off, and the screen was actually damaged from the heat. This was while the screen was down and it was still on.

I think that a laptop without any damage prevention could get to 120C -130C easy, probably even 150C. When you think about it, a desktop processor will only turn on for about 3 seconds without a heatsink or thermal compound, because there is no way to distribute the heat. As you can imagine from that, the processor will get extremely hot at an extremely quick rate.

 

Good post, my thought though is that there are absolutely no safety measures, triggers, airflow, or cooling in place. Just the chips heating up until they stopped getting hotter.

 

Well, assuming you want it to be 'on', it could hit 105C (or whatever the maximum TJunction temp of your CPU is) then the CPU would shut down. Or, much earlier, it would hit 55C or 60C depending on the HD you're using and essentially shut down. Or, you could sit in the Sahara desert at high noon and you'd melt and the computer would (for you) essentially shut down!

Curious if you're using your computer in an unusual 'hot' environment?

Cheers!

 

Well, as you are pure theory

Energy can't be "Lost" - so Energy in = Energy out

You have reaches the maximum possible temperature - considering no airflow or vents - once black-body radiation has reached the energy input

and its been a while since I last did physics...
Here:
http://galileo.phys.virginia.edu/classes/252/black_body_radiation.html

 

my media center had some cooling unit not on a chip. it could get up to 150° till the system froze.

and yes, pc's freeze when they have hot. quite stupid, not?

 

Oh! No safety measures at all? I think that silicone based circuits can survive up to 340C, but not during use. During use the electrical properties change and the CPU cannot be predicted any more - just random output and errors. I believe this temperature is around 120C.

 

Damn that's hot. For those that are wondering why I want to know it's simply just because, I don't use my computer in a hot environment it's simply my mind wandering and me attempting to further my knowledge.

 

Cool! Just another knowledge seeker!

I thought at first someone from the 'other side' was able to get a notebookreview.com account!


Cheers!

 

There comes a point where the solder will melt on the boards. BGA reflow temps are between 200-220 so I'd say anywhere in there and you've got a guaranteed dead notebook.

 

I've seen where Asus G50V's get up to 140°C.

 

Wouldn't the hardware damage prevention shut them down before they got to that heat level?

 

One would certainly hope so, after spending that much on such nice hardware!

 

Depends on what component's overheating triggers the shut down. If the laptop shuts down when the CPU gets above 105°C, the GPU can still get well above that as long as the CPU stays below 105°C.

 

The GPU also has its safeguards. It'll downclock as well.

 

Well if you're ignoring temperature shutdowns, a processor when started without a heatsink will destroy itself in seconds. The limit I'd guess is how far it will go before the silicon breaks down (which it does at a much lower temperature when energized than when not).

I used to have a watercooled desktop that I once started with some old crusty thermal paste on there (yeah, I was stupid. But it was highschool. That's my excuse) that was preventing ANY contact between my water block and the heat spreader. It barely got past POST before it shut off.

 

mine didn't destroy itself. it just turned off very quickly

 

safe max temps of a system vary depending on the part of the computer.

I think the gpu is the highest temp you will see if your on a gaming type laptop.

They can safely run 100c+ with there shutdown point in like 120c+ range, though most users start to feel uneasy after the 80-90c point and start to look into better cooling.

The cpu is in 2nd place, a modern core cpu can get into the 60's without problems and most people start seeking better cooling if they see it hit 70c+ with its thermal shutdown also somewhere in the 100c+ range.

Other parts of the computer can get quite hot, like the RAM and also the northbridge but these parts usually have no temperature probes on them because there is no real way to cause them to heat up beyond there normal range unlike the cpu/gpu that the temps range depending on load and if they are overclocked.

The important thing to know is the computer will throttle down to prevent damage to itself, this doesn't mean its fool proof but it does make it quite hard to damage a machine from overheating.

 

That's not a fair assumption. I once owned an HP dv6424ca with a Turion X2 TL-58 and an Nvidia Geforce Go 7200. The machine got so hot one time that it warped the plastic table it was resting on as well as the fan grille on the computer.

 

It could only get as hot as the CPU thermal shutdown limit is, for most MOBILE Core 2's for example it is 105C, it would also depend on what power plan you are on at startup, it may not even shut down when the thermal specification is reached, it usually just downclocks from what i heard, with the Pentium 4's that's what i know they do, same goes with GPU's.

 

i cook eggs on my laptop all the time. fans and heatsinks are overrated.

 

i would say max for CPU would be 105C but for GPU anything above that would do but modern GPU's will downclock... old ones like ATI X1000 series could hit 118C or something and still work till now...

 

Found the thread:
Asus g50v overheating

Looks like 140°C was the highest:
(from the thread)

 

Wow. I'm surprised the heat hadn't been transferred through the case to the CPU.

And also, minimum temperature for the CPU was -41C?

 

Yeah, that is kinda suspect those temps - I mean, not that they weren't accurately presented by the OP, just that the probes and/or software were not reporting the actual temps in the locations indicated.

I'm probably wrong, but I would trust a hardware temperature 'gun' more than the built-in probes on the MB/chipsets.

Cheers!

 

yes.. for example, TPfancontrol's GPU temps are always 10 degrees cooler than reported by HWmonitor

 

So... does that make HWmonitor more wrong?

 

Good lord.....that's awesome and completely not cool at the same time.

 

Those temps look a bit too high... must be altered....i doubt the GPU will reach 140C... even in gaming... would have downclokced by then and made game unplayable...so around 100C should be max...

 

My Gateway 7811 ran hot enough to melt the keys on the keyboard(i have picture proof). I think the maximum temperature before shutdown was 100C.

 

Do share my friend!

 

If you say so. Here is the thread I created about it a long time ago when it happend
http://forum.notebookreview.com/showthread.php?t=293778

And here is a picture(the picture from the prevous thread is gone )

 

Wow. That's pretty extreme. I have never seen a laptop get so hot that it could melt it's keys...

 

Really soft plastics... After I got a new keyboard, I experiment with the old keyboard keys and they melted quite easily.

 

Damn man that's hardcore, how did the rest of the computer fare?

 

seemed okay..

 

quite some macbooks had that problem. pre-unibody of course. the white plastic could melt and get ugly-yellow

 

Wow, I'm surprised the manufacturers let that pass. Did they discontinue the line when they found they were prone to that?

 

And risk losing their reputation and increased production cost?!!
No way...
Solution:
Let consumers suck it up until their warranties ended.

 

wow and wow @ the picture! like the other dude said, that's gangsta!

i believe that was the gateway model which had keys made out of butter... you know... so you could type faster

i'm no good at comedy.

 

just tell the consumer to replace it if they stay quiet, like they did for the ipod or iphone that started burning..

 

lol that has got to be extreme... my laptop reached 100C for CPU before and GPU reached 85C-90C... but my keyboard didn't melt... luckily.... but sometimes it makes me wonder why notebook manufacturers don't test out their notebook for heat and temperature and all before selling... it will save them a lot of problems and money with the warranty and replacement... and also save us from toasty notebooks...

 

^ more importantly:

1) customers are lured into buying the latest and the greatest so the company makes a good amount of additional $

2) having customers do the r&d or answering why the keyboards melt saves the company additional r&d money.

3) only a small percentage of users actually game for long periods of time so, although all machines may have melting keyboards, only few will actually see that happen.

corners are being cut everywhere but that does not mean the customers should get a faulty product. especially not one where the f'd keyboard melts like ice cream.

 

My 9800M GTX got up to 102C before it downclocked.

 

my desktop ati x1650 gpu got up to 128C.It stayed at 128C for 7 minutes but nothing happened.still working like the first thay as it is.

my pavilion notebook's 8600m gs got up to 105C.It didnt downclock or shutdown.

Ubuntu's graphics drivers say that(it was nvidia station x or something)the threshold temperture for 8600m gs is 118C.

I dont know but some gpu and cpu's are suprisingly durable.

ati x1000 series as I said can reach up to 128C,and at that heat they don't downclock or shotdown.It it hard to burn a cpu or gpu,really hard.they are durable.

 

My 7811's gpu runs much much hotter than the CPU.

 

Wow, all you guys have such hot running notebooks. The highest I've ever seen my 3650 is 65C, and the highest I've seen my CPU is 60C.

 

In reality I think that silicon components can take a lot more heat than most power users (those who are aware of these things - like us here on this board) believe. CPUs and GPUs (when designed well - the Nvidia 8600m not withstanding this) have a temperature throttling point designed to spare damage, and I can guarantee that this number is well below the point at which the silicon will suffer any damage.

It is true that cooler will theoretically last longer, but I doubt that anyone will keep a computer long enough to see a failure related to a CPU running at 100c.

What is more important to the average user is percieved heat. That is, the heat they can feel and observe at the exhaust vents and through the case.

 

Explosivpotato,

I don't think it is just theoretical. Every computer that has ever acted up on me had, in one way or another, been a problem with effectively getting the heat it produced out. BSOD's, random errors, programs working when the computer first turned on, but not even 10 minutes afterwards, etc. AMD's were the worst in this respect for me (they ran much, much hotter than the Intel's I had at the time).

As an example of running a cpu 'hot' I think this thread personifies it:

See:
http://forum.notebookreview.com/showthread.php?t=437686


Now, if I can get this computer to run at 800 Mhz, I'm not only lucky (sometimes it still clocks down to about 200 Mhz, even when limiting it to 800 Mhz speeds) - but almost happy about it too!

The 'perceived' heat from this notebook? Let's just say you don't need a space heater for the office with it on (when it was working 'properly'). Very, very hot.

Cheers!