Laptop liquid cooling?

Burbot

Member
Jun 26, 2004
58
0
0
How do you think, is it possible that notebook builders will start transitioning to liquid cooling? Radiator could be placed in the back of the screen, and coolant could be passed through hinges (quite possible with this style of hinge) to hot components inside. Most normal folks would not really care, but those who buy 12-pound "transportables" might consider this. Maybe that's how dual processor notebooks will be introduced?
 

RedsFanGeoff

Junior Member
Nov 14, 2004
14
0
0
I think you have some good points there. I would say no, for the reasons of cost and clunkiness of a cooling system. I think for us custom system builders we can find the patience and usability of a water cooled system. But that added weight and cost would shy away any major corporate laptop system builder. Also, so much money is being put to use in dual core processor technology which is going to be what all systems will be based on for increases performance. That factored in with bigger cache sizes, were all going to be able to increase performance without "needing" to increase clock speed which usually increases heat.

but yea, thats my ramble...
 

cirthix

Diamond Member
Aug 28, 2004
3,616
1
76
how about we start making laptops for bodybuilders instead. oh, wait, you want to!!!
 

f95toli

Golden Member
Nov 21, 2002
1,547
0
0
There are already water-ccoled laptops, unfortunately I don't remember who makes them (Toshiba?)
 

Burbot

Member
Jun 26, 2004
58
0
0
Also, so much money is being put to use in dual core processor technology which is going to be what all systems will be based on for increases performance. That factored in with bigger cache sizes, were all going to be able to increase performance without "needing" to increase clock speed which usually increases heat.
We will add more cores. Two cores = double the heat. No matter which way you pick, CPUs will get hotter.

But that added weight and cost would shy away any major corporate laptop system builder.
There were times when CPU heatsinks were unusual. Then there were times when CPU fans were unusual. Then there were times when high-perfomance copper CPU HSFs were unusual. Sure, no one likes this, but we are moving into domain of blast furnace hot CPUs, and we have to get that heat out somehow. Liquids work a lot better then air for moving heat out of enclosed areas.

how about we start making laptops for bodybuilders instead. oh, wait, you want to!!!
Hey, at least all those geeks would get a bit of muscle.
 

Cogman

Lifer
Sep 19, 2000
10,283
134
106
heres an interesting question, what about heat pipes? Generally they are lighter, and I have seen the movible metal joints that they could incorperate. and you could probibly use the extra room from an oled monitor to add fans....
 

f95toli

Golden Member
Nov 21, 2002
1,547
0
0
I was not Toshiba; it was Hitachi
Hitachi started selling a water-cooled laptop in 2002, so it has been around for some time.
llink to old news

Toshiba used to have a modell with heat-pipes, I am not sure wheter or not they have continued to use that system.
 

Tiamat

Lifer
Nov 25, 2003
14,068
5
71
I think the notebook manufacturers will probably think of a better ventilation path (more direct airflow into and out of the heat source(s)) and implement more heat pipes. Water cooling is still too risky for the normal user - Especially for a notebook that will be jolted around, dropped, turned upside-down, etc.
 

Machupo

Golden Member
Dec 15, 1999
1,535
0
76
www.overclockers-network.com
i'm pretty sure a lot of companies already use heatpipes in laptops -- that's why the fan can be in the corner of the laptop body while the cpu is closer to the center (at least, that's the case in my dimension 8600).

heatpipes propably stand a much better chance of rapid introduction than water cooling as you can still move the heat, but there's no need for pumps, water, etc...

maybe we'll see TEC aided heatpipe cooling? gotta get the power requirements waaaay down on TECs to see that though, battery life would go into the crapper
 

bsobel

Moderator Emeritus<br>Elite Member
Dec 9, 2001
13,346
0
0
We will add more cores. Two cores = double the heat. No matter which way you pick, CPUs will get

Wrong, two cores does not = double the heat.

Bill
 

NewBlackDak

Senior member
Sep 16, 2003
530
0
0
The first heatpipe I saw in a computer was a dell Latitude CP with a Pentium 200 MMX in it. The machine was about a year old when I saw it, and the model was about 6 months old when we bought them here at work. Heatpipe technology isn't new it's just been recently that it's been needed on a personal computer.
 

vegetation

Diamond Member
Feb 21, 2001
4,270
2
0
I don't see any need for it. My 1.6 dothan HP ZT3000 doesn't even need the fan on for anything but the most cpu intensive tasks.
 

Burbot

Member
Jun 26, 2004
58
0
0
Wrong, two cores does not = double the heat.
So two cores are not going to produce double the heat amount of a single identical core? How's that?
Now, one can sure downclock cores in a dual core package to bring the heat output down (just like Intel and AMD are going to do with their multicore processors), but if all you do is add second core, you get twice the heat.
 

w00t

Diamond Member
Nov 5, 2004
5,545
0
0
the mac G5 laptop is suppost to have water cooling i am not really into mac but i heard it from a mac guy
 

cheesehead

Lifer
Aug 11, 2000
10,079
0
0
Yep, the dualie 2.5 has watecooling.

Aside from this, I think a circulated heatpipe design would be spiffy. In other words, use a heatpipe, but put the radiator behind the screen and stick a very small pump in there somewhere to keep coolant moving. It would be low-noise, and it might cool a bit better than current designs.
 

Matthias99

Diamond Member
Oct 7, 2003
8,808
0
0
Originally posted by: Cheesehead
Yep, the dualie 2.5 has watecooling.

Aside from this, I think a circulated heatpipe design would be spiffy. In other words, use a heatpipe, but put the radiator behind the screen and stick a very small pump in there somewhere to keep coolant moving. It would be low-noise, and it might cool a bit better than current designs.

The whole point of a heatpipe is that it doesn't need a pump. If you're going to introduce a pump, you're creating a conventional watercooling system. Pumps also add size and weight, and create more heat of their own (so while your processor might be a little bit cooler, you'll have more total heat to remove from the chassis).

Heatpipes are also *widely* used in laptops these days, as well as SFF systems. They've been in laptops for a long while. It's just that, because they cost more than a standard heatsink with fan, they were unusual for desktop machines until fairly recently.

So two cores are not going to produce double the heat amount of a single identical core? How's that?
Now, one can sure downclock cores in a dual core package to bring the heat output down (just like Intel and AMD are going to do with their multicore processors), but if all you do is add second core, you get twice the heat.

Obviously, two identical cores at the same clockspeeds are going to produce twice as much heat as one core. However, heat output increases way faster than frequency does (it's basically proportional to V^2 * f, but leakage currents get worse as the frequency and temperature go up, and faster processors often need more voltage than slower ones to run stably); two processors running at half the speed and reduced voltage would likely put out half as much total heat or less when compared to a single faster processor. Compare the wattage ratings on (for instance) low-voltage mobile AthlonXP and Athlon64 processors, or the LV Pentium M versus the P4-M (or even the regular Pentium M). You get a lot less heat output per Mhz if you can step back from the bleeding edge a bit and lower the operating voltage.
 

JF0603

Member
Jun 27, 2004
44
0
0
i was wondering if the cpu got to a temurature cool enogh to melt liquid nitrogen it would cool the cpu perfectly. wouldnt it???
 

Burbot

Member
Jun 26, 2004
58
0
0
Obviously, two identical cores at the same clockspeeds are going to produce twice as much heat as one core. However, heat output increases way faster than frequency does (it's basically proportional to V^2 * f, but leakage currents get worse as the frequency and temperature go up, and faster processors often need more voltage than slower ones to run stably); two processors running at half the speed and reduced voltage would likely put out half as much total heat or less when compared to a single faster processor. Compare the wattage ratings on (for instance) low-voltage mobile AthlonXP and Athlon64 processors, or the LV Pentium M versus the P4-M (or even the regular Pentium M). You get a lot less heat output per Mhz if you can step back from the bleeding edge a bit and lower the operating voltage.
This is true, if you assume that second processor gives you 100% increase in performance. Unfortunately, dual core processors are limited by having only one memory controller. As a result, leakage decrease will be compensated by performance loss in shared bus system and limited application scalability, and performance/watt is not going to improve much (although responsiveness will be better).