Would a aftermarket cooler keep your room cooler or not? *poll*

[P4-630]

 

[EarthDog]

But do you understand what we are saying or just giving up and walking away thinking the same thing?

 

[silkstone]

again, high school level science here bud.

The confusion between temperature and heat is actually quite common.

 

[infrared]

The only reason for all this misunderstanding is people not being taught (or not remembering) any scientific fundamentals It's a shame really because a few basic facts and logical thinking can help you figure out a lot.

 

[EarthDog]

The confusion between temperature and heat is actually quite common.

indeed.. until highschool... for me. And trust me, I'm not the sharpest knife in the drawer.

 

[Air]

I had no idea a thread like this could actually generate discussion. Arguments being used:

1. Arguments that don't care for energy conservation: please...

2. CPU operating at higher temperatures have higher power draw than on lower temperatures: That's true but i bet i wont make any measurable difference in room temperatures, assuming like +5 W. I dont know the exact numbers though.

3. Thermal inertia of the heat sink will impact the room temperature (the one with graphs): While true this effect is negligible. It will not take hours before coolers reach their final temperature. I don't have a aftermarket cooler, but if its anything like GPUs, it will take more like 5 minutes after starting the CPU load... after that, heat output will be the same regardless of cooler. Not only that, but if you later stop the CPU load, CPU/heat sink temperatures will go back to idle levels, unloading the stored energy back to the ambient.

So, in the end, if you run your CPU load non stop, it will indeed keep a little bit of heat out of the ambient, but when you are generating a lot more heat for hours, it doesnt matter. If you run a variable load, it does not matter since on lower loads the heatsink unloads the stored heat to the ambient.

 

[CAPSLOCKSTUCK]

unsubbed.

 

[Kelvin]

In most cases the room will get hotter, as aftermarket coolers are often used for overclocking.
A CPU with higher speed & voltage will drain more power.

I remember a hot summer when I decided to undervolt & underclock my CPU & GPU to reduce the room temperature while gaming.
I got around 100W power consumption reduction and almost all of it went into reduced heat as microelectronics, I guess, are very thermal inefficient.
Although with current Intel CPUs the thermal efficiency tends to increase when you overclock, except if you decide to go for the last few MHz by going overboard with vcore.

In comparison an adult emits heat at around 100W.

 

[silkstone]

So, in the end, if you run your CPU load non stop, it will indeed keep a little bit of heat out of the ambient, but when you are generating a lot more heat for hours, it doesnt matter.

SHC of aluminium is 897 J/kg.K

Assuming a 1 kg heatsink vs. a zero mass heatsink, it will delay a 150 W cpu heating the room up by ~6 seconds.

Edit: (per Celsius difference in temperature between the two heatsinks) - A poor example as pointed out by Air.

In practice the room will get hotter, as aftermarket coolers are often used for overclocking.
CPU with higher speed & voltage will drain more power.

I remember a hot summer when I decided to undervolt & underclock my CPU & GPU (~100W reduction in heat as microelectronics, I guess, are very thermal inefficient) to reduce the room temperature while gaming.
In comparison a human emits heat at around 100W.

If the CPU is drawing more power, then more heat pumped out. Though I believe the initial question was just about the size of the heatsink.

The only thing I can think of why a bigger HS would make the room hotter is if it allowed turbo-boosting for longer and more frequently. That again implies a higher power draw,

 

[EarthDog]

I was wondering when the equation would come out.. .thanks silk.. hopefully that quiets this "10" who aren't understanding it. Also, there should be an equation which shows how much energy is needed to heat a given object (water is what I recall the equation being)... that equation doesn't mention temperature in it, except the goal temp you want to heat the medium to..

 

[Assimilator]

The fact that almost 10% of people of a technology forum don't understand basic physics does not bode well for the future of our species.

 

[silkstone]

I was wondering when the equation would come out.. .thanks silk.. hopefully that quiets this "10" who aren't understanding it. Also, there should be an equation which shows how much energy is needed to heat a given object (water is what I recall the equation being)... that equation doesn't mention temperature in it, except the goal temp you want to heat the medium to..

You're likely thinking of the equation for SHC, it tells us the amount of internal energy that an object holds at any given temperature.

My calculations above is actually not entirely correct as there is more at play and you don't get a zero mass heatsink

 

[Air]

SHC of aluminium is 897 J/kg.K

Assuming a 1 kg heatsink vs. a zero mass heatsink, it will delay a 150 W cpu heating the room up by ~6 seconds.

That's if the temperature delta is 1 °C... which wont be the case. You should use 30 - 40 °C, and keep in mind the heatsink will be losig energy during this heat up period, so i will take a lot longer. Not hours, but not 6 seconds.

 

[silkstone]

Thinking about it, the amount of heat (energy) that both a large heatsink and small heatsink will hold is likely to be similar due to the lower temp of the larger one.

Here's some calculations. I can work in Celsius vs. Kelvin as I'm just looking at the temp difference. The numbers I just pulled out of my backside.

Small heatsink (0.25kg @ 80 degrees ) The amount of energy it will hold is simply 0.25 x 900 x 80 = 18, 000 J

Versus

Large Heatsink (1 kg @ 50 degrees) Q = 900 x 50 = 45 000 J

Difference = 27 000 J

if it's being heated at a rate of 150 W (J/s) then it will take 27 000 / 150 = 180 s = 3 minutes longer to heat up (120 seconds vs. 300 seconds)

Edited for clarity

Edit @Air - Yes an over simplification of a complex system. If you want to consider everything, you'd also need to consider Boltzmann's law as well as a plethora of other things (That if I fully understood, I'd be getting paid a hell of a lot more). The larger heatsink would actually begin pumping heat out to the surroundings at a much higher rate when the computer is first turned on due to the surface area and convection. It's take longer to reach its equilibrium temperature, not because of the heat capacity, but because of the surface area.

 

[Air]

You're likely thinking of the equation for SHC, it tells us the amount of internal energy that an object holds at any given temperature.

Actually its the internal energy difference between 2 different temperatures. Not absolute.

 

[RejZoR]

Bottom line is, it doesn't matter how big the cooler is, it's how hot the source is. The rest is just a matter of time. If the cooler is massive with slow fans, it'll just release the heat slower. If the heatsink is tiny with fast fan, it'll blow tons of heat into the room. Over time, both will heat up the room to same temperature, it's just a matter of how long would it take for both to reach the same temperature point.

Good analogy for that are room heaters. If you have compact 2kW fan heater and massive 2kW oil radiator, they'll both consume same amount of energy to heat up same volume of air to a specific temperature. They'll just need different time to achieve that. This is essentially the same thing as for example CPU with 2 different coolers. The source will always be the same I don't know, 130W TDP CPU. It'll always release 130W of heat (lets leave out the power saving tech and no load situations to make it easier). When you generate 130W of heat, it has to go somewhere. And it'll go somewhere...

 

[Vulcansheart]

Bottom line is, it doesn't matter how big the cooler is, it's how hot the source is. The rest is just a matter of time. If the cooler is massive with slow fans, it'll just release the heat slower. If the heatsink is tiny with fast fan, it'll blow tons of heat into the room. Over time, both will heat up the room to same temperature, it's just a matter of how long would it take for both to reach the same temperature point.

Good analogy for that are room heaters. If you have compact 2kW fan heater and massive 2kW oil radiator, they'll both consume same amount of energy to heat up same volume of air to a specific temperature. They'll just need different time to achieve that. This is essentially the same thing as for example CPU with 2 different coolers. The source will always be the same I don't know, 130W TDP CPU. It'll always release 130W of heat (lets leave out the power saving tech and no load situations to make it easier). When you generate 130W of heat, it has to go somewhere. And it'll go somewhere...

OMG thank you. Finally someone that gets it. It's the rate of change from Tmin to Tmax that counts here

Edit
And I appreciate those of you that can propose an argument without slinging insults around like a child.

 

[silkstone]

Bottom line is, it doesn't matter how big the cooler is, it's how hot the source is. The rest is just a matter of time. If the cooler is massive with slow fans, it'll just release the heat slower. If the heatsink is tiny with fast fan, it'll blow tons of heat into the room. Over time, both will heat up the room to same temperature, it's just a matter of how long would it take for both to reach the same temperature point.

Good analogy for that are room heaters. If you have compact 2kW fan heater and massive 2kW oil radiator, they'll both consume same amount of energy to heat up same volume of air to a specific temperature. They'll just need different time to achieve that. This is essentially the same thing as for example CPU with 2 different coolers. The source will always be the same I don't know, 130W TDP CPU. It'll always release 130W of heat (lets leave out the power saving tech and no load situations to make it easier). When you generate 130W of heat, it has to go somewhere. And it'll go somewhere...

More or less. It'd take a different amount of time to distribute the heat rather than heating up quicker or slower, but that would mainly be down to the size of the fans.

Also remember that a computer heatsink is in a metal box. If you put a fan and oil heater in 2 different metal boxes, there would be no difference. They'd both heat the room the same

OMG thank you. Finally someone that gets it. It's the rate of change from Tmin to Tmax that counts here

Edit
And I appreciate those of you that can propose an argument without slinging insults around like a child.

Power is already measure of the rate of change of heat. If the power of 2 systems is the same the rate of change of temperature will be the same more or less

 

[Air]

OMG thank you. Finally someone that gets it. It's the rate of change from Tmin to Tmax that counts here

I get what you are saying, but I was trying to point out that both the difference in heat exchange rate before "Tmax" is reached, and total time before "Tmax", are too low to actually matter. After Tmax is reached, heat exchange will be the same regardless of cooler. And after the CPU load ceases, there will be another period of time before "Tmin" is again reached, where the oposite happens, the bigger cooler dissipates more heat.

 

[Sasqui]

The fact that almost 10% of people of a technology forum don't understand basic physics does not bode well for the future of our species.

Nor the fact that this thread is even still alive! W-T-F ?

 

[dorsetknob]

" In the name of Science"
i decided to Experiment
i set up 4 coolers of vastly different size some were just plain Aluminum and some were Aluminum with a copper core

temps were measured and checked (23c) Ambient room temp
all units were run for 30 min
There was no change in the Temp of the Room ( or in the Coolers ) after 30 min

conclusion small or big cooler "it made no difference to the Room Temp "

Ps i did not use a CPU as that was a variable i did not want to account for

 

[Steevo]

Same within the tolerance allowed by error.

If you have an enclosed space of a 10 meters and have 200 watts of heat dissipated into that area, the only thing changing a cooler may do is to change the overall localized hot spot, but the heat flux will remain the same, the only variance should be how much heat can the enclosed area dissipate in a given time, and the latent effects of heat soak.

Years of running water cooling after years of running air cooling, sound understanding of cooling and heating, and running overclocked Pentium 4 systems in enclosed spaces tell me X watts in means X watts of heat generated irregardless of the cooler attached, the only change being the CPU may run a few degrees cooler until the room temperature reaches its maximum.

 

[Vulcansheart]

I get what you are saying, but I was trying to point out that both the difference in heat exchange rate before "Tmax" is reached, and total time before "Tmax", are too low to actually matter. After Tmax is reached, heat exchange will be the same regardless of cooler. And after the CPU load ceases, there will be another period of time before "Tmin" is again reached, where the oposite happens, the bigger cooler dissipates more heat.

That's exactly what I was trying to demonstrate with my diagram.

 

[rruff]

CPU operating at higher temperatures have higher power draw than on lower temperatures: That's true but i bet i wont make any measurable difference in room temperatures, assuming like +5 W. I dont know the exact numbers though.

For example my 5970 dropped power draw by almost 20watt going from 90c to 40c under water. But this is negated by heat generated by water pump and ect.
In the end every PC is fancy electric heater.

This is the only thing of interest regarding this question. If the CPU is dissipating the same power, then that heat is going into the room one way or another.

But... if the CPU produces less heat if it's cooler, that may result in slightly lower room temperatures, provided that the cooler itself doesn't generate so much heat that it negates the savings. Power draw is a good measure. If your cooling system draws 20W more than your original, and the CPU draws 20W less, then nothing changed as far as the amount of heat dumped into the room.

 

[R-T-B]

Well, unless any thermodynamic engineers step forward to settle this, I think it's going to be idiots vs dumbasses for the life of this thread.

I'm as close as you're likely to get and frankly, no, this is high school level science. It's really sad.