Air Cooling with Heatpipes,deadend?
Ive seen more and more radical ways of heatpipe coolling this last year and I've got to say that is quit impressive.Ive seen everything from the Scythe Kamikaze to ThermalRight's XP120.As Cpu's advance (and OverClocking becomes even more popular)are there going to be any way to keep these beast's cool with heatpipes and fans or is it a deadend?Will new pc case's start coming with more radical ways to cool? Will watercooling become an industry standard? These are all questions Ive been asking myself lately and Ive been doing alot of research into it.Ide like to hear other people's input on this.Please Post back with your thoughts and research results.
0
Comments
Heatpipes are actually quite effective and if you are looking for a good solution look into the Zalman CNPs9500 it's brand new and reviews seem to be quite positive check them out here @ www.zalmanusa.com and a review I read on it was on www.pcstats.com. I have used heatpipes on video cards and they do an awesome job as a quiet cost effective and low-noise solution also for media centre's they do a great job, watercooling is not always an option for people as you are aware of the cost. but thermaltake and gigabyte for example make cheap internal watercooling for around $160 canadian so the trend is going there, Koolance also makes a very nice/pricey internal watercooling unit you can view that @ www.koolance.com in terms of cooling overall cooling though I doubt anyone would disagree that Zalman is one of the better and cheaper 3rd party heatsink manufacturers they also tend to make their product quiet.
Though right now the trend is - faster cpu's eat up more voltage and so give off more heat.
And I imagine Intel has thought about the heat associated with packing two cores into the same space as one, in an already thermally inefficient processor. Perhaps they really will use the Pentium M as their primary core, some people seem to really like it.
But back to my story. It was a while ago now so most people would have heard or read it, but AMD has a patent for a new method of cooling.
See, problem was that AMD's new method of cramming more porccesor onto the same die (Silicon-on-Insulator) generates a lot more heat. Basically because the silicon-oxide on which the silicon lies (they're sort of like islands in it) doesn't conduct heat to the same degree as the silicon itself, (we're talking 1/2 to a 1/3 the efficiency) AMD proposed (and patented) an on-die peltier system.
See, extreme.
Also if you look of late into some commercially built systems (my mate's Dell, specifically) you'll see an interesting inclusion, heatpiping and fruckin huge HS's.
His machine is a top of the line kinda beast, and has a 120mm exhaust (which, ingeniously enough doubles as his CPU fan), so it needs it to stay icy in such a restrictive case, but the point shows. Business are starting to move more hardcore too.
For example, back in the day of celerons, (the original ones) aftermarket HSF's usually consisted of a little hunk of metal with 1 or 2 30mm fans blowing over it.
TWO?!?! Why the hell use two? why not something bigger? It's because the amounts of metal and sheer size of the fans in modern heatsinks would have been completely unprecedented.
Now, celerons were a long time ago, but as CPU's get hotter, I feel that air-cooling will again get bigger. Perhaps we'll finally see a stock cooler using an 80mm case fan.
At any rate, I doubt copper will be a popular material (for the whole HS at least) as the weight to efficiency ratio simply isn't high enough to justify it for the most part. There are also costs involved.
Although then again, the BTX standard may prove me wrong. Who really knows?
I don't like the idea of heatpipes, but I don't like it because I can't do the math, and get a good figure, Here, I'll go over it with you..
Est convection rate and energy/heat transferrance of aluminium is 2.5W/HU.
Est convection rate and energy/heat transferrance of copper is 1.14W/HU.
--Now, since most heatsinks sit ON the processor, they do hold alot of heat against it, but also apply moving air on it. So, let's take my Thermaltake TR2 into play, at 33.5CFM, screwed copper base and aluminium fins. Which means, My copper base is holding alot of heat, that the aluninium has to ditch.
--Therman resistance of metal, Higher the total, higher the heat resistance is--
2.5+1.14=3.65.
3.65/31=8.49 C of complete thermal resistance. That 8C is sitting on my processor, doing nothing.
Now, since heatpipes are ranged, they can be exposed to more air, but also have less surface area to transferr the heat.
Solid copper will be used for this, since I haven't seen an aluminium one.
1.14/31=27C. That makes NO sence, since in that case, the heat would never LEAVE the chip, and it would eventually overheat. Also, since heatpipes would radiate that heat throughout their path to anything, you're also heating up your case unessisarily.
So you see what I mean, The math for heatpipes doesn't make sence, because if it's supposed to be good, why does it leave that much heat behind? It can't possibly leave that behind and be an effective cooling method.
I have as stated above a Thermaltake TR2, which I paid 7 bucks for. My 64 3000+ has never been at 35c.
Anyway, that's just my opinion, I could be wrong.