"Mass-market budget water-cooling" - Is it driving users away from water-cooling?

[HAL-9000]

Noise and price are what I would say is the barrier to mass-acceptance of water-cooling. If I buy a nice Swiftech or Thermalright sink, with appropriate low-noise fan, I'm out about seventy bucks. I cannot put together a water-cooling system for that price that will beat the air solution in either performance or noise. So there really is no point.

Even with a nice setup, you don't see a big improvement in temps until you compare temps when the computer is working hard. A good aircooler can maintain virtually identical idle temps as a nice watercooler. Factor in the reliability, simplicity, and aesthetics of a good aircooler (the new ones are quite pretty) there is no reason to go water until you hit $150 and up in your cooling budget. So that precludes about %70 of the enthusiast market right there, forget about Grandma.

But, aircooling is getting quite extreme in design and application just to be a competent cooler on today's procs. Watching CPU's go from no sink to all copper 80mm x 80mm HSF for stock Dell cooling is a testament to this. Video cards are going through that right now. Video card aircooling is going to get very problematic in the next 24 months for two reasons:

1.There are actual market forces and applications that push vidcards year after year...and that market is willing to pay big premiums for performance. MS Excel in Office 97 is pretty much MS Excel in Office 2K3 horsepower-wise. Doom 1 is nothing like Doom 3, so the vidcard makers are going to continually push harder on performance, even if CPU's slow down in development.

2. Peripheral card form-factors are not very accomodating to continued aircooler growth. You will not see a Swiftech 6400-V with monster fan bolted onto a Hercules Prophet unless user of such a setup is willing to forego having any other peripheral card in his rig.

I think the PC industry is watching what happens with Apple's watercooling "experiment." If user-feedback does not reveal big problems, I think you will see some other OEM's take the dip.

[Tempus]

One other factor is the maintenance or "dust bunny X-factor."

I work in enterprise IT and I occassionally let users bring home PCs in to get some help (we offer that as incentive to get them to do the home-office thing.) Most of these people have older PCs (P3\500-P3\800s being the norm right now.) On average, these cases have never been cracked. There is a LOAD of dust bunnies gumming things up. Most of the intake and exhust fans are absolutely coated. I've had to replace cpu HSF for the same reason - they burned the motors out due to the dust and grime.

Fast forward to a more modern machine after 3-5 yrs of dust with no cleaning. What will happen once those nice copper heatsinks get full of insulating dust and the fans stop spinning so well because a 3" dust bunny jumped in there. Not to mention the front intake grill is completely covered over with pet hair and dust.

This is where watercooling COULD make a deeper penetration.

W/Cing is not ever going to be as cheap as a $70 HSF, regardless of the economy of scales being brought to bear. But, we should be able to see something under $150 if/when major names start playing around and shipping 1000s of units.

Noise and reliability will drive consumers to demand a change from traditional air cooling. Modern processors (both cpu and gpu) will be far less forgiving in a dusty home.

I don't think you will ever see a day when mass-market KITS become a big item. Mass-market people do NOT want to open their computers, much less remove and add things to the cpu.

High-end enthusiasts will continue to migrate at an increasing rate. One big reason is the trend to SFF. With SFF, the heat density increases very quickly, fresh air (and unobstructed airways) becomes harder to get, and the coolness factor of WC will all combine to make it the premier choice. That day is probably 18-24 months away - less if Nvidia's SLI takes hold and the extra heat load really starts to push the serious games into WCing. Have you see how close those cards are? Talk about no air flow.

/rant

[HAL-9000]

Originally Posted by Tempus

One other factor is the maintenance or "dust bunny X-factor."

I work in enterprise IT and I occassionally let users bring home PCs in to get some help (we offer that as incentive to get them to do the home-office thing.) Most of these people have older PCs (P3\500-P3\800s being the norm right now.) On average, these cases have never been cracked. There is a LOAD of dust bunnies gumming things up. Most of the intake and exhust fans are absolutely coated. I've had to replace cpu HSF for the same reason - they burned the motors out due to the dust and grime.

Fast forward to a more modern machine after 3-5 yrs of dust with no cleaning. What will happen once those nice copper heatsinks get full of insulating dust and the fans stop spinning so well because a 3" dust bunny jumped in there. Not to mention the front intake grill is completely covered over with pet hair and dust.

This is where watercooling COULD make a deeper penetration.

W/Cing is not ever going to be as cheap as a $70 HSF, regardless of the economy of scales being brought to bear. But, we should be able to see something under $150 if/when major names start playing around and shipping 1000s of units.

Noise and reliability will drive consumers to demand a change from traditional air cooling. Modern processors (both cpu and gpu) will be far less forgiving in a dusty home.

I don't think you will ever see a day when mass-market KITS become a big item. Mass-market people do NOT want to open their computers, much less remove and add things to the cpu.

High-end enthusiasts will continue to migrate at an increasing rate. One big reason is the trend to SFF. With SFF, the heat density increases very quickly, fresh air (and unobstructed airways) becomes harder to get, and the coolness factor of WC will all combine to make it the premier choice. That day is probably 18-24 months away - less if Nvidia's SLI takes hold and the extra heat load really starts to push the serious games into WCing. Have you see how close those cards are? Talk about no air flow.

/rant

I know what you're talking about. Back in the dark days of my phone-tech support career I supported HP Pavillions. There was actually a fail code in the Peoplesoft database for repairs that was listed for "infestation." Sometimes cockroaches would move in because it was warm and dark. Cockroaches gum up fans and CD-ROMs when they hop on for a spin....eeeew.

[jlrii]

Maybe the engineers @ Dell are sitting there calculating the eventual death from dust bunnies to coincide with hardware obsolescence. LOL

[Cathar]

Originally Posted by AngryAlpaca

Nice. Very nice. I hadn't thought about the higher flow. Haha. This testing would be extremely helpful for pushing forth custom setups, gives me some credibility. I don't think, for this purpose, that we should use a D4 or a 120.3, as they aren't really favourites. Double heater core and a Mag 3 (obviously we'd fix this before sending)? Using really high end, expensive parts won't help in the least as people'd see the massive price tag and say, "Screw it, I'm going to go with an Aquagate and save me $150"

AA, you need to consider a few things:

1) Volume Pricing
2) Quality components - I don't subscribe to the naff budget pump approach - maybe I'm weird, but if with volume discounts we can get reasonable quality pumps into people's systems, then that should be done. The experience should not be wrought with a high pump failure return rate with people asking "Why is my system beeping at me?"
3) Nice looking components - a skanky looking heater-core that looks like it just got ripped from a car wreck just isn't going to cut it.

With volume pricing, and packaged in a neat external case with a knob to turn the fans down in volume, we should be easily talking about a $200 price-point with those components.

Have the radiator and pump in an external case, where users can easily see it and see what the radiator is doing with respect to dust buildup on the external face of the radiator, should be good enough for most to get the hint to vacuum the radiator each time they vacuum their floors.

[AngryAlpaca]

I didn't realize that it was actually a product to sell. I thought this was just showing the superiority of the method... In that case I'd say a C-systems pump as the D4 would probably turn people off of the product due to the noise (one of the main advantages of watercooling is to get rid of the whiny fan). I never thought that a brand new double heater core looked too bad... A little bit of spray paint with a proper shrould would fix it right up.

I think $200 might be pushing it just a little, unless we're in insane volumes. 120.3's are going to be expensive unless we're in the hundreds of thousands... profits to procooling?

[Etacovda]

I think this is a possible product, not an end product...

The question is the ability to provide a low maintenance, high reliablity set up without compromising on parts, and still providing top end performance.

I think its shifted somewhat now though, to 'how would we show top end cooling well on joes testbed to guage these kits properly vs whats theoretically possible with top end cooling?'

Seems like a few trains of thought now.

[jaydee]

Originally Posted by Etacovda

I think this is a possible product, not an end product...

The question is the ability to provide a low maintenance, high reliablity set up without compromising on parts, and still providing top end performance.

There are many questions before that one gets answered. Like "is it worth our time and money for R&D?".

If something like this gets done it will not consist of any parts we currently know of being they will all be specially designed and fabricated for their systems like the new Mac one. Lots and lots of cash will need to be burned before a product will even hit "prototype" stage.

[firtol88]

Originally Posted by Etacovda

...The question is the ability to provide a low maintenance, high reliablity set up without compromising on parts, and still providing top end performance...

Build it and they will come.

Originally Posted by Etacovda

...I think its shifted somewhat now though, to 'how would we show top end cooling well on joes testbed to guage these kits properly vs whats theoretically possible with top end cooling?'

Seems like a few trains of thought now.

What he said.

I still stand by my pick for a reference point.

[HAL-9000]

I always thought a good product evolution for mainstream watercooling would be an integration of the pump and the radiator.

It would work something like this: You would have a common motor for both the radiator fan and the pump. A small plastic gearbox would adjust the RPM difference between fan and impeller for optimum efficiency of both. The pump (and motor) itself would be housed in the dead-space of the fan's hub. A spiraled tubular"ring" would cascade out of the pump through concentric copper fins directly behind the fan blades. This would provide a radiator with no right angles to flow, reducing resistance within the radiator. Inlet would be at the front, outlet would loop back around through the fan housing of the gadget. When I think of this thing, I think of a unit sized for a 120mm fan housing.

The advantage would be if you did this, and integrated the actual resevoir into a compartment either as part of the radiator/pump or the waterblock; it would reduce the "kit" to two components. The 120mm form factor would make it ready to go into anything ranging from a 120mm fan housing to a replacement for a Shuttle XPC heat-pipe setup. OEM's like Alienware would dig it because it would be potentially higher performane, and far simpler, than a Koolance. All kinds of ancilliary costs would be reduced. You only need one housing, not two. You need only one motor, not two, one pair of tubing, not two pairs...etc. If you wanted performance, you could crank the motor, increasing both the pumping capacity and the radiator's fan speed with a simple potentiometer. Just an idea...

[firtol88]

Production cost will not go down, the specialized one off parts would cost more than most existing systems.

[Colt357tw]

Cather has a good point about how cheapie watercooling kits gave a bad rep for watercooling and personally I cannot agree more. It is just so hard to make a great waterblock (or 2, or 3), radiator, pump, with fans and the hoses into 1 mid/full tower case without look like products would be used in Dr. Frankenstein’s Lab.....(woven steel re-enforced fuel hoses are really fun to play, but to make AN connectors for them are pain....)

[BillA]

Colt357tw

you posted a link to a spreadsheet
what is the x axis ? (cannot see the bottom)
who did the tests ?
other data ?
very interesting

[Colt357tw]

Originally Posted by unregistered

Colt357tw

you posted a link to a spreadsheet
what is the x axis ? (cannot see the bottom)
who did the tests ?
other data ?
very interesting

please see PM

[travanx]

I would be a decent example of someone new to watercooling. I basically read a review about the Kingwin kit. And thought wow prices finally dropped this low on a WC kit? Maybe I will finally look into getting one. I wasn't really looking at it as something that would let me OC any higher though. More like possibly a better solution for noise and possibly cooling the chip better to make it more stable. My AMD64 runs very hot when its OC'd, though it is stable.

So I looked more into the kits and noticed they weren't very good after doing some message board reading. I finally got the idea of what made a basic/decent kit with quality parts. Its a little more expensive than what I originally read about, but it wasn't enough to make me back away from the price. $250 shipped from danger den for a kit to cool my Radeon 9800pro and AMD64 3000+.

Now the two things that made me reluctant all this time to try WC. One was the reliability. 2nd was maintenance. But it seems like a lot of kinks have been sorted out of this cooling method. Still not perfect where anyone could do it. But its getting very close.

Bigger stores carrying the quality parts and more information on the parts seems like the main things needed to get WC into the general market.

I am into cars, so when Subaru brought out their WRX with a turbo, suddenly cheap little turbo cars started selling fast. But Subaru didn't price the WRX cheap and use cheap parts. They priced the car moderately and made it reliable. Now look at how many little cars are coming out with turbos. Of course the car analogy mainly pertains to the US market.