watercooling lost its magic

[Jag]

QUOTE=Marci]
Market saturation hasn't occured YET, but for the kind of watercooling to which we refer (the extreme enthusiast / DIY end), it more or less has.

Over-Clock.com is churning out 40% of what it was a year ago. We used to be shipping 50+ custom kits out a day. Now we're lucky if we move 1 full kit a day.[/quote]

Marci,

Agree on the "saturation" issue, BUT isn't it related to the simplistic vision of "pure performance", leaving all other aspects behind?
Two years ago everyone was in a "race" to have the best performing waterblock.
The truth is that very few people see a WCing setup as a reliable cooling solution, and therefore have a lot of resistance towards its use.
As BillA so eloquently wrote (#26 - http://www.ocforums.com/showthread.php?t=476949)and quote: "the aggregate MTBF of 3 fans will always exceed by far any WCing system".
Isn't this the real issue STILL not adressed by anyone - enthusiast and manufacturers alike?

All of a sudden some users here just opened their eyes and realised that aircooling is the better option when not so long ago, these same people were marvelled and even went as far as buying mills.
Is this logical?

[migueld]

I do agree there are many other aspects of W/C to be exploited and it is up to the enthusiasts to lead the way.

[Brians256]

I see HTPCs and NAS boxes a lot like watercooling and overclocking was 2 years ago.

[DVL73]

Originally Posted by Jag

The truth is that very few people see a WCing setup as a reliable cooling solution, and therefore have a lot of resistance towards its use.
As BillA so eloquently wrote and quote: "the aggregate MTBF of 3 fans will always exceed by far any WCing system".

I simply don't understand why people are moaning about the WC stability and especially adding the MTBF in equasion. I've seen so many cases of cheapo fans dying on the CPU heatsink (mostly because of the dust and quality issues) and messing up the system. GPU sector is another story too.

Self contained, properly built, low maintenance WC loop will last for a years to come and what everyone is forgetting is that because of lower WC average temperatures of other system components (CPU/GPU) they will last longer and their MTBF is improved. How many cases of the DDC faults we have seen, if we exclude the usual user faults ? Are your cars safe and reliable ? cooling of the engine is based on the same principle.

Personally, I don't see much practical sense in BillA statement.

Originally Posted by Jag

All of a sudden some users here just opened their eyes and realised that aircooling is the better option when not so long ago, these same people were marvelled and even went as far as buying mills.
Is this logical?

Logic is simple. Passion and excitement is gone. At some point in your life, I guess that you are more concentrated on the practical things and as Marci said things that are more important in life ... It's very individual, but on the other hand it's perfectly normal.

It's not important if someone throw the hands up in the air and give up. Important thing is what they done and what they left behind. That may include the inspiration too, for the youngsters.

[Jag]

Originally Posted by DVL73

I simply don't understand why people are moaning about the WC stability and especially adding the MTBF in equasion. I've seen so many cases of cheapo fans dying on the CPU heatsink (mostly because of the dust and quality issues) and messing up the system. GPU sector is another story too...

Do you have a WCing setup in your computer?Guess not.

Originally Posted by DVL73

Self contained, properly built, low maintenance WC loop will last for a years to come and what everyone is forgetting is that because of lower WC average temperatures of other system components (CPU/GPU) they will last longer and their MTBF is improved. How many cases of the DDC faults we have seen, if we exclude the usual user faults ? Are your cars safe and reliable ? cooling of the engine is based on the same principle..

Does Apple G5 ring a bell?

Originally Posted by DVL73

Personally, I don't see much practical sense in BillA statement..

Well, BillA's statement is exactly in the same context as the others, so you seem to be contradicting yourself, defending some and condoning others.

Originally Posted by DVL73

Logic is simple. Passion and excitement is gone. At some point in your life, I guess that you are more concentrated on the practical things and as Marci said things that are more important in life ... It's very individual, but on the other hand it's perfectly normal.

It's not important if someone throw the hands up in the air and give up. Important thing is what they done and what they left behind. That may include the inspiration too, for the youngsters.

[DVL73]

Originally Posted by Jag

Do you have a WCing setup in your computer?Guess not.

I do have, why is that important now and more importantly why such attitude

If you want to take a look:

http://forums.procooling.com/vbb/showthread.php?t=13439

Originally Posted by Jag

Does Apple G5 ring a bell?

Yes, it does. I was just mentioning global concept and didn't want to quote any particular brand.

Originally Posted by Jag

Well, BillA's statement is exactly in the same context as the others, so you seem to be contradicting yourself, defending some and condoning others.

You didn't got the point. I was just commenting the "eloquent" BillA statement about comparing the 3xAC/fans MTBF with the WC loop. Obviously, you did like it very much.

[Joe]

increased MTBF on a vid card or CPU just from cooling? Whens the last time you had a video card, or CPU just die cause they wore out? Ive seen vid cards die from bad caps, or other failures, but never just die caue they were warm. the MTBF on the silicon components is quite high... higher than almost anything else in a PC.

A good CPU fan and sink will last maintenance free for many years... and if it does fail... you replace it with another 10$ fan. Almost every CPU/Motherboard has thermal shutdown thresholds, and protection in the CPU's against firey death if a fan fails on a HSF.

Pretty much the only place I could see H2O cooling increasing MTBF would be on HD's. The cooler you can keep a hot drive ( a 15K SCSI for example) the better spindle life you will get.

[migueld]

Originally Posted by DVL73

Self contained, properly built, low maintenance WC loop will last for a years to come and what everyone is forgetting is that because of lower WC average temperatures of other system components (CPU/GPU) they will last longer and their MTBF is improved. How many cases of the DDC faults we have seen, if we exclude the usual user faults ? Are your cars safe and reliable ? cooling of the engine is based on the same principle.

While I do agree that a properly built w/c loop will last quite some time, I don't think the reliability is as good as air cooling yet. But this is just another challenge to take upon.

Originally Posted by Jag

Does Apple G5 ring a bell?

The Apple G5 was a first generation product and Apple gave up on liquid cooling too soon. Their new solution is hardly better so I wouldn’t consider that as a good example.

[DVL73]

Originally Posted by Joe

increased MTBF on a vid card or CPU just from cooling? Whens the last time you had a video card, or CPU just die cause they wore out?

Heat does degrade and I think in a measurable way too, apart from the fact that I really don't have the definitive formula right now. Heat will kill your CPU or GPU eventually and the more heat, the quicker.

In some general perspective, it should be our objective that we run our systems at the lowest "possible" temperatures and it doesn't matter if we OC or not. I'm really not sure about the definitive "proof" how the CPU/GPU MTBF life coincides with the practical life, influenced by the temps ... but we do make an impact on that. Maybe additional question is what is our tolerance on how hot our components run.

Many moons ago, I killed one Radeon 9700 card because of the heat and one ancient P4 CPU. It was OC-ed, yes, and my air cooling couldn't cope. Actually it was coping on the edge. Each of the component died after approx month of usage. After WC upgrade, I OC-ed a LOT and didn't have any single failure - ever. I don't have the great airflow in my case (and didn't want fans and nosie) and WC helped a great deal as all other air cooling solutions are really more dependable on the case ambient air then WC itself.

Originally Posted by Joe

A good CPU fan and sink will last maintenance free for many years... and if it does fail... you replace it with another 10$ fan.

Correct and I can't argue with that (and make sure that fan has ball bearings and even better that rotor is sealed inside of the casing) ... but what we are doing in this forum section then?

Originally Posted by Joe

Almost every CPU/Motherboard has thermal shutdown thresholds, and protection in the CPU's against firey death if a fan fails on a HSF.

But ... this is not air cooling advantage, if I understood you correctly.

Originally Posted by Joe

Pretty much the only place I could see H2O cooling increasing MTBF would be on HD's. The cooler you can keep a hot drive ( a 15K SCSI for example) the better spindle life you will get.

There you go ... and HD is important device, isn't it


Honestly Joe, maybe we are getting more and more practical. When practical solutions and thoughts are in the zenith, magic is gone for good and this is maybe answer to your question. What's wrong with the little bit of magic nowadays, even if it's not practical or little bit DéjÃ* vu ?

[DVL73]

Originally Posted by migueld

While I do agree that a properly built w/c loop will last quite some time, I don't think the reliability is as good as air cooling yet. But this is just another challenge to take upon.

Thank you migueld I think that we are finally going to the the hard core of the matter: are we there yet ?

[jaydee]

Originally Posted by DVL73

Thank you migueld I think that we are finally going to the the hard core of the matter: are we there yet ?

Give someone, able to do anything about it, a reason for it to be....

[pH(x)]

Originally Posted by Joe

OK I Posted the editorial Enjoy!

http://www.procooling.com/index.php?...icles&disp=148

Brilliant work Joe, I didn't have time to read the whole thread here. But I certainly read that

I am the owner of a system that got me a 700MHz overclock back in 2002, now this machine is just a shadow of it's former self -- it has fallen into disrepair for a lack of human enthusiasm

[Errorz]

Many good topics have been spoken about.

I must agree with those that wonder of those stating the reliability of watercooling. My previous watercooling experiment is over 5 years old (the coolant from my first system, 7-8 years old), I dont use it everyday, its had ample time to sit around and rot quite a bit, never done any maitenance on it... Still chugging fine. Reliability, in my opinion, is fine. Plastics and tubing have never degraded for me.

I must agree with most of everyones comments as to why watercooling has lost its alure. For the past couple years I have been on a laptop and perfectly fine.

Though I must remember why I watercooled in the first place. I too started back when dangerden and dtekcustoms didn't even exist, started with overclockwatercool for me (oh mayn! it took me 5 minutes trying to remember it!).. its for the fun of it all. There were great performance gains back in the day, but sound and "coolness" still always availled. The impossibility of trying to explain watercooling to those eager eyes It's for the fun of it all!

Right now I am thinking up a small custom mini rig. There is something about mixing just the right % of coolant, turning the pump on for the first time, watching the water cycle, letting it sit for a day or two for leaks(or often times just saying screw it and jumping in), tinkering with tubing, seeing how quiet you can make the fans and still get away with it... This is why I still wanna do it... It's fun! I hope we can look at the differences in market while still remembering what it is that brought us all into the hobby... It's fun! MUCH more fun then simply plugging everything in and shutting the case ya?

(for those wondering who I am, I am what you call the occational lurker of 7 years, when procooling didn't have a forum hehe... )

Have a great day

[Pyrotechnic]

I kind of feel the same way about the whole deal. I'm typing this on a watercooled system though.

I remember when I got into watercooling...back in middle school. Back then I had lots of time, and enough money saved up to really play around with computers and water cooling. I thought it was so bad ass to be able to use water to cool the CPU, keep everything quiet, and get a nice overclock. So I did. I must have done about 10 revisions and used 2 different cases before I settled on the system I'm still using today. It all started from stuff bought from home depot mixed with plumbers goop, to a well engineered and designed full featured system with little compromise thats still giving me reliable cooling to this day.

Today is different though. I'm 240 miles from home living in a dorm room. I dont really have the tools or resources, money, or time to even mess with a watercooling system anymore. I barely even have time to practice drums or bass, or play with my band back in my hometown. I guess I dont have time for a lot of the "magic" things that I liked to do.

I think it all goes around in cycles though. The 486's and Pentium's ran really cool. When that technology was pushed, it kept getting hotter and hotter, until they changed the manufacturing process, and things have cooled down again. Software will become more demanding and the limits of hardware will be pushed again.

Either way, I'm still running a Socket A chip and it's watercooled. I love how quiet and how reliable the system is and i'll continue to use it as long as I am able to. The only maintenance involved is blowing the dust out of the radiator, and once a year, changing out the coolant for good measure, although it has lasted much longer in the past.

[Edge]

Yeah, I came in to water cooling at the end of the magic. I had a Tyan S2460 dual socket A with 60mm howlers on the heatsinks - looking for quiet pushed me to passive radiator water cooling.

This past year I was doing well with air cooling, but what goes around comes around - the cpu's aren't putting out that much heat, but the dual cores/quad cores are cooking the chipsets (which have really annoying noisey tiny fans) and the graphics cards have just gone over the edge of stupid on heat.

Watercooling, lost its magic, now it's just plumbing to deal with the components that aren't keeping up with the processors (bait).

[Anemone]

Remember too, that for many the degree of improvement has changed. This was touched on earlier in the comments about just how much cpu power is "needed" these days. 5-10 years ago the difference in power between a low run of the mill machine and a top powered one, was power that some folks really used.

What I find is that it's really the classic, how much money and effort for how much improvement formula at work. Years ago many of you saw this coming. If I have a heatsink with a reasonably quiet fan that comes within 5-10C of the same performance, and within 200mhz of oc'ing performance of the watercooled system, what then is the return for spending on the watercooling system? Most folks are not going to notice much in 200mhz, and frankly silicon can stand 5c and often 10c higher temps for years without any real concerns as long as it falls within spec. That tends to make for a next to zero return, for more money and more effort. That's nice if there is something that really gives you satisfaction in building the thing, or watching it run, or perhaps the additional silence it provides, but most folks aren't going to bother.

Air cooling only needs to come to 90% of the ability of water cooling and then cost, effort to install, and performance return on the cpu take over. And don't compare a poor air cooling system to a well designed water cooling system. Apply equal amounts of design to both and you'll find that the pinnacle of each is extremely close in performance. The case and airflow, noise and fan choice, all come into play, but when equal amounts of good thought are put into both kinds of designs and maximized for your choice of performance or noise, things can come pretty close. And close is fine, because many of us are edging out all they can from a cpu. But cost does come into it. Spend $200-300 on a cooling system, when in a year or two that same money will buy you a cpu that is vastly faster than the increase you got out of the current cpu? The math, the very "need" for that extra few mhz of performance, just don't support a broader market acceptance. All this at a time when laptops are the largest growing sector of the pc market. People want quiet, surely, but they want smaller, more portable, the whole "chic factor" more. And if the laptop market is an example, they are happily willing to pay more and give up performance at the same time! Funny as it seems even high end air cooling feels the pain. Let alone a market where the costs are at least double air cooling.

Sorry this got too long. But I think you all will see the idea here. The market for this sort of thing isn't nobody. There are enthusiasts aplenty. But the broad applicability for the technology has run into the other competitive choices and a degree of improvement factor, that are making things very difficult for water cooling.

[lloydrmc]

Originally Posted by Azazel

I just made an attempt in an LC17 from silverstone. Unfortunately, the pump was just too loud (DDC 18v version), even when enclosed in a metal box filled with foam and undervolted. A lot of compromises end up getting made, but it can be done. I just wasn't willing to put up with the noise.

For reference, it was a TDX, MAZE 4 GPU, Black Ice Pro II and DDC, all with tygon 3/8in. I'm pissed off that it failed, as now I've got all this WC gear and nothing to do with it.

Interesting that the editorial should mention the Arctic 64 Pro, because that is the HSF that has pretty much driven me to water cooling. I guess, despite my efforts to the contrary, I've managed to purchase a case with bad airflow and/or ruin the airflow, and the 64 Pro hasn't helped. In fact, when I put the side cover on the case, my poor Opty 170 overheats in less than two minutes.

I'm sick of fooling with fans and sick of the noise. I'm coming up with a quiet system, I'm putting water blocks on EVERYTHING (including replacing the heatsinks on the PWMs which currently have small, noisy, OEM fans on them), and I'm NOT going to spend anything in the same zip code as $180, even if that means I have to fab every last water block myself.

And, no, I'm not going to suck it up and pay some exhorbitant amount on relabeled fountain pumps (and blow my budget) - I have a line on a surplus place that has the real deal for a good price.

Maybe this stuff is old hat to you guys because you've done it so many times before. I want a my bigger OC out of my Opty to process BOINC work units faster, and I want it to be done quietly.

You know, that's the "killer app" as far as I'm concerned - your CPU is at essentially 100% load essentially 100% of the time, and if you don't think that taxes your cooling, then think again. Especially if you run an Opty like I do.

I am frankly underwhelmed by most of what I've seen on the retail market - there seems to be a lot of "we have a big reputation and our product has a cool name, so pay $150 for this radiator that's not demonstrably better than a $30 heater core"

Now, I most notably exclude a new product that I'm aware of that was engineered in large part with input and discussions from some very knowledgable people right here on this board - such product designed to work with the kind of fans we use, not some noisy squirrel-cage blower whirring away under the dashboard of a Chevette or whatever. I would consider spending the money on one of those radiators, because it really is such, and has demonstrable performance commensurate with the asking price.

[Brians256]

The last pump that I used was a Laing D4 (not a DDC) and it was louder than my current setup which is air.

I use air cooling because it is quieter in my application. For lower power computers, it is hard to beat air cooling. This is because the pump noise is louder than the equivalent fans. You still need air movement in either case. Water is used to spread the heat over a larger area so that the air velocity to remove the heat is lower. That is all.

For an overclocked or high-heat CPU such as yours, water cooling can far exceed the abilities of air-cooling favorites such as Scythe HSFs with 120mm Nexus fans. My CPU is only 65W and it's only that high when it's fully loaded (not for more than an hour or two).

Frankly, I don't think there is a good argument here. Water cooling is more appropriate in cases such as yours, but air cooling happens to better fit most people's situation.

[Heavy_Equipment]

I read the editorial when Joe posted it, and I agreed with the main message on the direction of the hobby, even though I was water-cooling at the time.

My water-cooling experience started as an OC tool, back in the Celeron 300a days, and lasted through the Athlon XP 2500+. All in the name of getting more for less.
In the end, it was out of necessity. Efficient heat exchange was key, since I had a DC herd with 12 rigs.

My final kick at the water-cooled can was this:


One big pump, a 5 gallon res, 2 heater cores and fans hooked up to a dryer vent to exhaust the heat outside. I even set it up in front of an AC unit.
It never evolved past the 6 rigs hooked up to the loop as seen here, but it sure was fun to build.
The last of that DC herd, and any subsequent water-cooling went offline about 15 months ago.

Fast forward to present, and here I sit with a Q6600 Quad with stock HSF, OC'd to 3GHz, chewing through Rosetta@home, and I can hear the wall clock ticking away beside me. 12 GHz in one very quiet and moderately cool box.

I have lost any need/want/desire to squeeze any more out of this system, and it's easily the quietest PC I've had in 10 years. My $270 Q6600 is doing $1100 QX6850 speeds right out of the box, so why bother?

Water-cooling served it's purpose for me, but at this stage of the game, it's now water under the bridge.

[ben333]

yeah but put that quad on water and go from the total of 12 GHz to 16 GHz or more! Thats why I am going to water cool my quad with an air conditioner compressor.

[Heavy_Equipment]

I converted a dehumidifier to a water chiller once upon a time, but my priorities have changed.
For me, the returns don't justify the investment anymore.

[HammerSandwich]

Looking at power vs. performance, I think an E8400 upgrade is my best pick. If only I could find a CPU for sale... The Wolfdale should be low power, and aircooling whispers to me in the dark of night.

[ben333]

no. Air is evil. I can't wait to get my quad wet again. I'm running only at 2.6 right now

[lloydrmc]

Originally Posted by Brians256

The last pump that I used was a Laing D4 (not a DDC) and it was louder than my current setup which is air.

I use air cooling because it is quieter in my application. For lower power computers, it is hard to beat air cooling. This is because the pump noise is louder than the equivalent fans. You still need air movement in either case. Water is used to spread the heat over a larger area so that the air velocity to remove the heat is lower. That is all.

I keep seeing statements like this, and I think they completely ignore the fact that water conducts heat better than air does.

Originally Posted by Brians256

For an overclocked or high-heat CPU such as yours,

try BOTH!

Originally Posted by Brians256

water cooling can far exceed the abilities of air-cooling favorites such as Scythe HSFs with 120mm Nexus fans. My CPU is only 65W and it's only that high when it's fully loaded (not for more than an hour or two).

Frankly, I don't think there is a good argument here. Water cooling is more appropriate in cases such as yours, but air cooling happens to better fit most people's situation.

I'm not trying to argue. I've seen where people have built perfectly good rigs with air cooling. I've never managed to. My latest case has a system temp of only about 9C higher than ambient, which would indicate to me that air is moving in the case, and the CPU still runs way too hot for my taste.

I think I can acoustically isolate the pump better than I can an air-cooled case. Worst case I'll just put it in another room entirely

BTW My Opty has a TDP of 110W, and my BOINC projects certainly push it right up against that.