Intel new cooler ?

[joemac]

http://www.xbitlabs.com/news/cpu/dis...920001532.html


I can’t help but wonder how many dust bunnies will be born from such a creature. I think Intel (and AMD ) are being a lot like ford was, resisting water-cooling until it could no longer cool it’s engine just by air. Nothing wrong with that but why keep beating a dead horse when there is a healthy one standing next to it.

[Zhentar]

because its not a dead horse, there is still quite a bit of life in air cooling, if you're not going for a great overclock quiet air cooling still has a great future.

Not the least of problems with watercooling would be the increased cost.

[joemac]

Zhentar
because its not a dead horse, there is still quite a bit of life in air cooling, if you're not going for a great overclock quiet air cooling still has a great future.

Not the least of problems with watercooling would be the increased cost.

So the future of cooling using current technology is air-cooling? I doubt that, I have been getting comments from users that have purchase my block stating that not only are there temperatures down drastically but so is the noise level. Isn’t this what the CPU manufactures are looking for? Cost is not an issues because they don’t sell the CPU with cooler but they do recommend the minimum required to cool.

[Joe]

Aircooling still has a load of life in it.

Sorry but water cooling will NEVER be the primary cooling method for PC's.

just a fact.

[joemac]

joe
Aircooling still has a load of life in it.

Sorry but water cooling will NEVER be the primary cooling method for PC's.

just a fact.

say it anit so Didn't people once say that a heat sink and fan was excessive. I am not going to argue that air cooling still has a long life but it’s nice to know that there is an alternative to 60db noise and barely cutting it cooling.

[Zhentar]

Originally posted by joemac
it’s nice to know that there is an alternative to 60db noise and barely cutting it cooling.

Yeah, there are lots of alternatives... lots of people at SPCR are cooling high power CPUs like the 3.0C sufficiently with less than 25 db noise. Unfortunatly, getting a computer to that point can be quite expensive and time consuming.

For example, the current quietest stock PSU is the Seasonic Super Tornado 300W at $70, and if you need more power than that (which you almost never do) You can go with the slightly louder Seasonic Super Silencer 400W for $100. And even then many people want it quieter than that, For example the 350W fanless PSU at a mere $350

Oh yeah one very big other problem for watercooling mainstream- Joe Blow is not going to think his OEM pouring water in his computer is leet. Every idiot knows water ruins computers!

[DigitalPirate]

my $0.02
I have to agree with joemac. The new Dells are all using heatpipes, which are (liquid) in a sealed tube. If a company like Dell is doing this, perhaps watercooling is the future. Consider what we have seen these past few years. Watercooling went from a completely DIY project to prebuilt kits. A minimal WC setup isn't that much more expensive than those top-end aircoolers. Sure, WC is still really expensive, and the average computer user doesn't give a D!CK how it works, just that it works. I should know, working as a PC tech. If they can get the same reliability out of WC as air, price will be the only consideration. And it's true, CPUs are getting hotter. The heat load is increased, therefore a more efficient solution is called for. Water has a heat transfer coefficient that is much better than air. If you want to use a wind tunnel to cool your CPU, great. I do acknowledge that in the end, you have to dump the heat somewhere, thanks to conservation of energy, and air is still the place to do that, seeing as we have so much. But right on the CPU? Why not try something new? WC in general, and joemac in particular seem to be focused on the OC'er. Air cooling still has a home with the average user. Feel free to flame me now.

[Pritorian]

Well i love watercooling, but as DigiPi said, Dell uses Heatpipes, and as thing progress i would think that most OEM´S will go over to this kind of coolinf solution.

Just look at the lates Zalman Case, totally no fans at all, and everything is cooled by heatpipes and larger heatsinks..

I would think that aircooling will survive the long run and water/phase change etc will still be considered leet

Only diffrence i could see is that aircooling solution are gonna get bigger inn weight and size, so that the surface area is bigger then what todays solutions are.

[Joe]

I agree heatpipes are to come, but water kits with any amount of water is just not happening for mainstream and primary cooling. User A who just wants a high speed machine doesnt want to have to add water to it...

[jaydee]

Originally posted by Joe
I agree heatpipes are to come, but water kits with any amount of water is just not happening for mainstream and primary cooling. User A who just wants a high speed machine doesnt want to have to add water to it...

What he said. There is no need for loud air cooling on even the fastest retail CPU amd or Intel has right now. My little XP1700+@2000mhz (A core) 2.0V full D.F. load (about as much wattage as the fastest chip out there) runs fine with a Volcano 6cu+ with the fan turned down to where you can hear the hard drive clicking. Thats over 500mhz on air to boot.... Heat pipes are going to be an alternative but they still need to "remove" the heat somehow. All the heatpipe does is move the heat from the CPU to the other end of the pipe where it needs some kind of active cooling to compete with aircoolers. At least thats how they look to me. Water cooling at this day and age is just way to impractice and way to much maintenance. Image trying to keep up 2500+ water cooled computers at large business.

[maxim]

i believe that we will see OEM WC if the CPU's keep producing more and more heat. it is possible that a new technology could be used to make the CPU's cooler, but if it doesnt, then we'll have to resort to better cooling. what the WC kit for a regular user will be like is something like this:

• closed loop
• all-in-one kit that includes WB, rad, fan, small but powerfull 12v pump (solid state maybe?)
• short hoses that are permanently attached and sealed with the rad and WB
• new case designs could be used to make installation of these kits easier

imagine having the standard case designed with a special window cut out on the side to mount a little radiator+fan in. the hoses that run just far enough to the WB with the right amount of slack, because all CPU's are in the same area in these cases. no need to refill or do any maintenance, except maybe blowing out some dust bunnies from the radiator. this can also be replaced with a system that uses an air filter that can be cleaned easier (think clothes dryer).

[joemac]

Originally posted by maxim
i believe that we will see OEM WC if the CPU's keep producing more and more heat. it is possible that a new technology could be used to make the CPU's cooler, but if it doesnt, then we'll have to resort to better cooling. what the WC kit for a regular user will be like is something like this:

• closed loop
• all-in-one kit that includes WB, rad, fan, small but powerfull 12v pump (solid state maybe?)
• short hoses that are permanently attached and sealed with the rad and WB
• new case designs could be used to make installation of these kits easier

imagine having the standard case designed with a special window cut out on the side to mount a little radiator+fan in. the hoses that run just far enough to the WB with the right amount of slack, because all CPU's are in the same area in these cases. no need to refill or do any maintenance, except maybe blowing out some dust bunnies from the radiator. this can also be replaced with a system that uses an air filter that can be cleaned easier (think clothes dryer).

I think you are right – Most OEM wont aim for 5 degree’s above ambient. Instead they will aim for temps with in spec’s. How do I know this? I use to work for an international company that sells business machines. So the temp that OEM will aim for is about 135 F – so why not use air-cooling some may ask? Because the power (Wattage) being put out by the newer CPU will be higher where regular air cooling won’t cut it.

[jaydee]

Originally posted by joemac
so why not use air-cooling some may ask? Because the power (Wattage) being put out by the newer CPU will be higher where regular air cooling won’t cut it.

I doubt it. A CPU will have to get around 150+watts with "current" aircoolers in order for this. AirCoolers still have room for advancement to boot. I belive they will start redesigning CPU's to put off less heat long before water cooling will become needed. The OEM's will DEMAND it IMO. Not much maintenance in a aircooler. Change a fan every couple years. As opposed to more serious problems water cooling brings up. Leaks, failed pumps, ect.. And you STILL need a fan on the rad. Remember your STILL removing just as much heat as an aircooler. Just because it is water cooled doesn't mean the CPU magically puts off less heat.

I think we will see aircoolers with larger (less noisy) fans on an aircooler instead of that same fan on a rad with a bunch of other possibilites for failure.

I also think if more cooling is needed than air cooling can provide then they will skip water all together and go phase change. You can almost put together a phase change system from a $100 Wal Mart Window AC unit cheaper than you can a decent water cooling system.

[DigitalPirate]

Alright jaydee,
Let me ask you this. What do you suppose happens since CPUs continue to decrease in size? Heatsinks can only spread a over a limited area, known to me as the heat bloom. This is why all the good WBs have such thin bases, right? Even with new heatsink technologies, copper has an upper limit on the amount of heat that it can pass at ambient temps. So either we move to different materials for heatsinks, which surpass the saturation limit of copper at ambient, we supercool the copper, increasing the amount of thermal differential, or we move the heat away from the source, allowing it to be dissipated over a far larger area at ambient temps, like a radiator. Before I start sounding like a college professor, I'll say I don't like lecturing, since by your other posts you have been around the waterblock several times.
I reject your conclusion that phase-change systems will be de rigueur anytime soon. They still have too many issues, such as the possibility of leaks(easily solved), component failure, and condensation(easily fixed). The possibility of component failure is far too great, though. Compressors will become quieter, but they may stop working, or jam, and oops, bye-bye CPU. Before phase-change can be mass-marketed, it must be silent, like heatpipes. If joe customer has the choice between a supercooled PC and the same one that is silent, he's going to take the quiet one. I would in his place, and I've seen enough other people do the same thing. By the way, I run a Xeon DP system that happens to be watercooled right now, but am waiting for Cathar to try fitting a couple of his beautiful Cascades to Xeon mounts. So I am partial to WC, as it has not let me down yet. Yes, any active component can break, and as Murphy's Law tells us, the possibility of a piece breaking is inversely proportional to it's importance.

[jaydee]

Originally posted by DigitalPirate
Alright jaydee,
Let me ask you this. What do you suppose happens since CPUs continue to decrease in size? Heatsinks can only spread a over a limited area, known to me as the heat bloom. This is why all the good WBs have such thin bases, right?

No not even close, but this is why the best AIR coolers have such a THICK base.

Even with new heatsink technologies, copper has an upper limit on the amount of heat that it can pass at ambient temps. So either we move to different materials for heatsinks, which surpass the saturation limit of copper at ambient, we supercool the copper, increasing the amount of thermal differential, or we move the heat away from the source, allowing it to be dissipated over a far larger area at ambient temps, like a radiator.

The upper limit has not even come close. Also there IS new materials coming out that is twice as conductive as copper. See other thread HERE that YOU started . Move the heat away from the source to be disipated over a larger area? Heat pipe wold be better at this and less chance of failure. No water, no pumps, no leaks, no maintenance. Phase change system WILL dramatically get better and cheaper over the next few years. And you thinking to small. Once phase change unit could cool many computers and be stored in its own sound deadened room. But this is still not much of an option. Also do not forget about the heat spreaders on the newer CPU from AMD and Intel.

Before I start sounding like a college professor, I'll say I don't like lecturing, since by your other posts you have been around the waterblock several times.

Ok

I reject your conclusion that phase-change systems will be de rigueur anytime soon. They still have too many issues, such as the possibility of leaks(easily solved), component failure, and condensation(easily fixed). The possibility of component failure is far too great, though.

not any greater than water cooling. Frig's last how longthese days? Air conditioners? They don;t break nearly as much as water cooling gear. The biggest failure of the system is the pump.

Compressors will become quieter, but they may stop working, or jam, and oops, bye-bye CPU.

Never heard of C.O.P.?

Before phase-change can be mass-marketed, it must be silent, like heatpipes. If joe customer has the choice between a supercooled PC and the same one that is silent, he's going to take the quiet one.

Water cooling these HOT CPU's is NOT, I repeat NOT going to be silent. The more heat involved the the more air your going to have to push through that rad, the more air to push the louder the fan. Water cooling is not a answer to a silent computer.

I would in his place, and I've seen enough other people do the same thing. By the way, I run a Xeon DP system that happens to be watercooled right now, but am waiting for Cathar to try fitting a couple of his beautiful Cascades to Xeon mounts. So I am partial to WC, as it has not let me down yet. Yes, any active component can break, and as Murphy's Law tells us, the possibility of a piece breaking is inversely proportional to it's importance. [/b]

I have 3 water cooled systems and over 60 waterblocks I have made over the last 3 years and know the advantages of water cooling and the dis advantages.

Also do not forget the more moving parts you add the more chance somthing will break. Air coolers have one breakable part that can be changed in minutes. Large companies are NOT going to settle for water cooling because of price and cost of maintenance. It just isn't going to happen.

With advances in thermal materials and advanced in CPU design that will increase power and lower wattage I see nothing strong to indicate water cooling will go mainstream. It is simply to impracticle. There will be better options that have yet to even be though up IMO.

I could be wrong though. And I definatly CAN wait to find out being this really will not be an issue for years to come. I will still be using a XP system in 3 years and will start considering a Athlon 64 around then when 64bit software slowly starts creeping out.

[DigitalPirate]

I deserved that, and no, I haven't heard of C.O.P.
Otherwise, you have made your point...
I think I'll shut up now.

[jaydee]

Originally posted by DigitalPirate
I deserved that, and no, I haven't heard of C.O.P.
Otherwise, you have made your point...
I think I'll shut up now.

C.O.P. is "CPU Overheat Protection". C.O.P. is actually a Asus name but other mainboards are following along with their own versions. Intel is already a little ahead with the built in throtteling in the P4's. i hear the Opterons have similar code built in that may or may not be used yet.

Article: http://www.asusemag.com.tw/tech/ch5/ch5-1.htm


My ASUS A7V8X-X and my ABIT KD7 have this on them as did my ASUS A7V333. I think it is standard on all higher quality boards but I could be wrong.

Also not saying your completely wrong either, but I see nothing to indicate water cooling will be a viable option. I had this same argument 3 years ago and had the same opinion then as I do now. Just my opinion remember...

[Joe]

you guys to remember back in 1998 they were predicting CPU's by 2004 would be generating so much heat and eating so much power every machine would be water cooled. Since no one knew that advancements in core design would come this far to lower heat.

Just to point out that industry experts cant predict the future any better than anyone else. So really maybe if everyone just stopped trying to assume what the performance will be heat wise or core design wise in the years to come, we will be better off

Its all just FUD at this point cause really no one knows wtf they are talking about when it concerns the future of CPU design. who knows what breakthrough, or change in design that intel, ibm, motorola, etc.. come up with tomorrow, or the day after.

[iroc409]

i'd have to agree with joe and jaydee.

i think there's a lot of variables here that just can't be known.

the improvements in processor construction will likely reduce the heat output of the device. if i recall, the newer, much faster xp processors put out less heat than older models? i think this will continue to improve, and i doubt the increases will be as dramatic as some may think.

also, there will most likely be vast improvements in aircooling technology, as was pointed out. i mean, looking at the air cooling device i use now, and comparing it to the aircooling device i commonly used a few years back where i used to work, are QUITE different.

materials are another big part of that equation. not that long ago pretty much everything was aluminum for air cooling, but now that has advanced so a good margin is now using some part of copper to increase efficiency.

not only that, but you'll see design enhancement in the physical design of the devices as well that will improve performance dramatically as well, just as they have in the past.

and, i think as well the use of heatpipes will increase and become more efficient just as standard aircooling, and be an increasingly viable option. they will progress with everything else.

just my .02

[pHaestus]

I suspect you'll see more and more motherboard advances in temperature-based fan control. This is the simplest place to add complexity. Use a larger louder fan in case exhaust and on heatsink but allow it to be throttled down in noise when system is idle. We are starting to see this in some newer motherboards but I will boldly predict it will become much more important. MOST people don't keep CPU at 100% load all the time and so this kind of solution is smarter and cheaper than elaborate quiet cooling to accommodate the few that do.

No one is mentioning GPUs and honestly they are a bigger problem. There's just not enough real estate to REALLY cool 50+W down between AGP and top PCI. Some rather elaborate solutions (Abit OTES, Zalman heatpipe thing, etc) are being released but will these be enough in a few years if GPUs are pushing 70W?

[Joe]

Similar cooling issues were seen usign SECC Slot1 CPU's. Goign to a flat socket chip let more cooling be used easier. Wonder when an interface to just drop a PCB holding GPU, Ram, and some other minor parts onto the mobo in a socket form. let the mobo do all the interface work.

[prb123]

Besides if we move to quantum computing, a method of removing heat may not be necessary...or move to another device (possibly a laser)

Note: No one currently knows how a quantum computer will work in the future, but current research suggest light/laser, or magnetic fields.

Note 2: This will NOT happen for a long time, but with everything being thrown back and forth I felt like throwing a monkeywrench out there.

[nuclear]

mostly directed at phaestus and joe
Didn't intel introduced a new form factor to solve that problem ?
the cpu and the pci express card are now at the bottom of the case and they are both cooled (as are the voltage controllers) by the air coming from the front of the case and exiting at the rear. So there will be more place to have bigger heatsing on the gpu

[Joe]

WTX was introduced a long time ago to solve that. But it didnt catch on

[nuclear]

it's not wtx i'm talking about, it's a new form factor that will be introduced at the same time as pci express