Do low CPU temps harm it?

[m25]

We all know that high temps can toast a CPU but what happens when idle temps (the ones that are near ambient) are as low a 12-14°C or lower? My room in winterr stays much around this when I am away and when the CPU is loaded they max @ 22°C.

 

[shinigamiX]

No worry about that, extreme overclockers often cool their CPU to -100C or below... 22C@Load is awesome for a normal system, I'm jealous.. :?

 

[jaguarskx]

It will be fine. Those are not extremely low temperatures.

I would only worry about extremely low temps, like if you decided to use liquid nitrogen to cool your CPU.

 

[Dahak]

No worries there neibour.At those temps you should actually extend the life of your cpu.Rule of thumb for processors,the cooler the better.Goodluck.

Dahak

AMD X2-4400+@2.4 S-939
EVGA NF4 SLI MB
2X EVGA 7800GT IN SLI
2X1GIG DDR IN DC MODE
WD300GIG HD
EXTREME 19IN.MONITOR 1280X1024
ACE 520WATT PSU
COOLERMASTER MINI R120

 

[bluntside]

Fear not, your temps are really low, wouldnt be worried about it, just when temps hit aroun the upper 80's or 90's, then you'd start having problems.

 

[TechnologyCoordinator]

Those temps won't hurt your processor. If the temps were REALLY low it could cause it to expand and contract too much and damage, but we're talking sub-zero temperatures.

However, one thing to worry about is humidity levels. Ever open up your front door in the middle of winter and notice the glass of the storm door is completly fogged up? I just worry about the humid cold air hitting something warm and then condensation occuring. But I don't really know how humidy, air, and temperature interact.

 

[cdonato]

Nah, lower is def better. Processor life goes up the cooler it is plus you can always OC more if the temps stay low.

 

[TechnologyCoordinator]

I just worry about the humid cold air hitting something warm and then condensation occuring. But I don't really know how humidy, air, and temperature interact.

Actually, after thinking about it, I think it is the warm humid air(in the house) hitting the colder door that causes the condensation (to form on the side of the door INSIDE the house). I still have no idea what the humidy situation would be and how it would affect your computer.

 

[Guest]

Warm air can contain more humidity. The warmer the solvent (air) the more you can have water in it(think of it like hot coffee, you can have all your sugar diluted in it, if you let it become cold, sugar will add up at the bottom).

Also think of humidity as water that became vapor, so when it touch a cold surface, it looses internal energy and fall back to it's liquid form(condensation), if the surface is really cold, it can even fall back to solid directly(deposition).
(My chemistry is a bit rusty and English is my second language people correct my analogies/terminology)

So when you cool your CPU in sub zero, humidity touching the surface will either condensate(water droplets) or to deposit(humidity become ice directly), that can cause short circuit on your motherboard or electronics component around it.

Oh and getting cold will cause it to contract, on such a small piece of metal/silicon I dont think it can cause any damage because the contraction/expansion is minimal, not talking about a bridge here!

 

[croc]

Braggart...

 

[chuckshissle]

That's good actually, assuming you can stand the cold inside your room. But you should prevent your room from getting that cold, you might catch pneumonia.

 

[audiovoodoo]

That's good actually, assuming you can stand the cold inside your room. But you should prevent your room from getting that cold, you might catch pneumonia.

Sounds like the sort of customer Intel wanted back in the Peeshot days

 

[weilin]

haha, maybe he has a prescott. Lets overclock the puppy.. the house is getting too cold. Next stop, 10ghz!

 

[kugi]

OVERCLOCK!!!!!!

 

[YO_KID37]

OVERCLOCK!!!!!!

I Second this motion. You need, No Must! Over clock! !Vive Le Overcompensation's!

 

[m25]

I know, it's prime stable undervolted to 1.2V but first got to get a good board because this crappy A8V-MX does not go beyond 224 FSB.

 

[Aragorn]

For rusty Chemistry and English not being your first language you were right on. I probably wouldn't have noticed that english was not your first language (but then I would have thought you a bit of a lazy typist, so I am glad you cleared that up).

 

[Guest]

Hehe thanks, I am also a lazy typist, which makes funny sentence sometimes =)

Off course Firefox 2.0 spell check helps me a great deal!

 

[WizardOZ]

As labbbby pointed out, condensation is related to relative atmospheric humidity, which in turn is related to ambient temperature. In temperate climate zones, relative humidity is pretty variable, depending on season, weather conditions, proximity to large bodies of water, and so on. In winter, most people heat their houses. Also, the colder the "natural" ambient temperature, the drier the air is.

But furnaces produce extremely dry air, which is why most are equipped with a humidifier. I have seen ranges of 25% to 40% relative humidity recommended in some furnace/humidifier manuals. The reasons given include, but are not limited to, human health and structural integrity of the house. Healthwise, the humidity makes breathing easier and maintains moist mucous membranes, which trap bacteria better - less colds. From a house structure perspective, it is not good if the lumber gets too dry, or so I am led to believe. High relative humidity levels can lead to condensation in a cooled system.

Overcooling a system can lead to problems due to differential shrinkage rates. At the same temperature, different materials shrink or expand at different rates. Metals vs plastics vs composites have different physical characteristics, which can cause serious physical strains at best. Differntial component shrinkage rates can lead to significant strain on the MoBo and components attached to it.

And this is before we get into the issue of condensation on the componts of an agressive coolling system. And where is this condensation on the tubing and cooling blocks going to go, I wonder.

 

[Aragorn]

All of that is true. And all of that applies to a phase change system, a peltier system, and even some water cooling systems.

The OP was asking weather very low ambient temps were bad for the CPU I believe the answer here is a definate no they are not. He is only talking abotu an ambient of 12-14 C. Since the inards of the machine are still above the ambient condensation is not an issue.

 

[WizardOZ]

Aragorn, your observation re OP's question is correct. Thanks for the reminder. And generally speaking, a low ambient temperature won't hurt the system. Indeed, this is one reason why serious air-conditioning is installed in many computer rooms.

However, a few points need to be kept in mind re excessively low ambient temperatures.

First, peltier and phase change systems are somewhat esoteric to bring up in this discussion. Liquid cooling systems have some interesting condensation issues which are almost never raised. As this is outside the scope of the discussion, I won't go into it.

What is within the scope of the discussion is the issue of local hot-spots in a system and the effects of differential thermal expansiopn coefficients of different materials. In my business, I have used seperatory funnels on occasion. These are made of glass, with Teflon stopcocks. If the ambient temperature gets too low, the Teflon shrinks lots more than the glass, leading to leaks. Very simple example of differential shrinkage at the same temperature,

Now, in a computer, we have several localised hot spots: the CPU, the Graphics card, RAM, and the chipset. All of these components will heat the MoBo in the immediate physical location of the slot or attachment point of the respective chips.

If the ambient temperature is low, it is very likely that portions of the MoBo will be significantly cooler than those in direct contact with hot components, possibly leading to physical strain due differential shrinkage and or expansion. Unless the OP has a really good ventillation system on the computer, this is a real concern. The temperature range specified by the OP is not likely to cause problems with the CPU, but overall effect of low temperature needs to be considered. Too much of a good thing scenario.

 

[orsino]

haha, maybe he has a prescott. Lets overclock the puppy.. the house is getting too cold. Next stop, 10ghz!

You know, I've pretty well had it with people slamming the Prescott. It is a superb CPU that I can count on to perform flawlessly, reliably and to keep my house perfectly toasty warm through the upcoming cold winter! :lol:

"Who Needs High Efficiency Furnaces When You Have A Prescott?"

 

[Aragorn]

I do not believe taht anything you put in your response is wrong, but why should the temerature diferences in his sytem be any worse than those in any other air cooled system? Shouldn't he have the same differences between the ambient temp in the room and his components as I have in mine. It's just that they are all lower, in this case he said his cpu temp only goes up to 22C that is only 8 degrees above his ambient of 14.

I guess basically what I am asking is why would a lower ambient temp make hotspots more of a concern when the hotspots are also that much lower?

 

[WizardOZ]

Wierd things can happen at both low and high temperatures. I am not saying they will happen, but one would be prudent to consider the possiblities.

Ultimately, my concerns and points are grounded in a conservative approach to dealing with hardware. My assumptions and concerns could very well be wrong and misguided. If so, I appologize, and my points can be disregarded.

I recall an overclocking test done by Tom's Hardware where they used liquid nitrogen as a coolant ( boiling point is: -195.79 °C, -320.42 °F) with really amazing results, i.e. a really extravagent top end for OC. What was interesting was that there was no discussion of how long the system was actually run, and no discussion of the condition of the MoBo and components during/after the experiment.

Hotspots may be lower as well as overall system temp, and the physical structure of the system may be able to cope just fine. As long as things are reasonably constant I agree that there shouldn't be any real problem. But I note that the OP did not specify what the ambient and system temps were when he was actually using the system. Somehow, I doubt that s/he is maintaing an ambient room temp of 12-14 C when actually present in the room. Significant variations in temperature cause rocks to flake apart. MoBos are a lot more delicate than stone is.

Sadly, I have had lots of experience with hardware problems and failures when operating under "stock default" conditions to make me extremely wary of "non-routine" or "enhanced perfromance option" situations and setups. Additionally, I am wary of glowing reviews based on performance differentials of less than 2% between different products. Especially when long-term stability issues are not addressed in any way.

If it ain't broke don't fix it is a very good appraoch, especially with computers. No reason not to experiment, but just like gambling, don't play with your grocery, rent or mortgage money. If you can't afford to repalce damaged components or systems, leave well enough alone. Even if money is no object, and one has lots to piss away, I can think of better uses for the cash - like supporting organizations like the Scouts/Guides or the Red Cross etc. Especially since the stock performance of current systems is amazing as it is, even compared to systems from 2 years ago. Never mind the systems old geezers like me worked on (Apple II's , IBM PCs and old mainframes with punch card compilers).

 

[maverick7]

the mach 2 phase change unit goes to -55 around with a conroe cpu and thats just fine for it.. it also means oc to the max