Water Cooling 101 - Should I or Why Bother ?

[JackNaylorPE]

Been a lot of questions on the board of late on water cooling and having spent lotta time looking at stickies around the web, I'm noticing lotta information outdated and some still not there. And when i go back and search the THG Forum for threads I remember, I'm finding them hard to find again. So thought I'd start a series of threads for peeps to discuss and share their thoughts on recent activities, thoughts, research and efforts.

NOTE: These threads are not intended to address the closed loop, ALC Units from Asetek and sold as the Corsair H series, Antec Kuhler series, etc. as goals here are WC systems that provide better performance and lower noise than air coolers

The topics I addressed so far include clickable links below:

Water Cooling 101 - Should I / Why Bother ?
Water Cooling 102 - One Loop or Multiple Loops ?
Water Cooling 103 - Radiators and Fans
Water Cooling 104 - Reservoirs
Water Cooling 105 - Water Blocks
Water Cooling 106 - Pumps
Water Cooling 107 - Tubing
Water Cooling 108 - Fittings
Water Cooling 109 - Gadgets and Doodads
Water Cooling 110 - Coolants
Water Cooling 111 - Filling, Draining and Maintenance
Water Cooling 112 - Cases
Water Cooling 113 - Sick Loops - *Warning* - Spouse May Claim "Justifiable Homicide"

Here's the THG sticky on the subject for ya convenience.

http://www.tomshardware.com/forum/277130-29-read-first-watercooling-sticky

 

[JackNaylorPE]

The water cooling world to my eyes has changed. Internal CPU heat issues have taken center stage; it's one thing to get heat away from the IHS .... it's another to get it to the IHS. Has water cooling lost some of it's effectiveness unless you are willing to undertake "delidding". Most of us will go ahead and do it anyway for the fun of it but is there any way to justify the cost based upon performance gained ? Just how much better performance do we get moving from one of the better air coolers (Phanteks PH-TC-14, Thermalright Silver Arrorw, Noctua DH-14) to a:

Closed / Open Loop System - CM Eisberg 240L Prestige or Swiftech H220
$130 - http://www.newegg.com/Product/Product.aspx?Item=N82E16835103184

$150 WC Kit - XSPC Raystorm 750 EX240 Extreme Universal CPU Water Cooling Kit (Plus tubing)
http://www.frozencpu.com/products/21235/ex-wat-270/XSPC_Raystorm_750_EX240_Extreme_Universal_CPU_Water_Cooling_Kit_New_Rev_4_Pump_Included_w_Free_Dead-Water.html?tl=g30c83s137

$500+ - Full Blown Custom loops

$1000+ Full Blown Custom Loops w/ rigid tubing, elbow fittings and lotsa gadgets and doodads

What did you do, what was ya reasoning and what did you get out of it ?

 

[crooked windows]

The water cooling world to my eyes has changed. Internal CPU heat issues have taken center stage; it's one thing to get heat away from the IHS .... it's another to get it to the IHS. Has water cooling lost some of it's effectiveness unless you are willing to undertake "delidding". Most of us will go ahead and do it anyway for the fun of it but is there any way to justify the cost based upon performance gained ? Just how much better performance do we get moving from one of the better air coolers (Phanteks PH-TC-14, Thermalright Silver Arrorw, Noctua DH-14) to a:

Closed / Open Loop System - CM Eisberg 240L Prestige or Swiftech H220
$130 - http://www.newegg.com/Product/Product.aspx?Item=N82E16835103184

$150 WC Kit - XSPC Raystorm 750 EX240 Extreme Universal CPU Water Cooling Kit (Plus tubing)
http://www.frozencpu.com/products/21235/ex-wat-270/XSPC_Raystorm_750_EX240_Extreme_Universal_CPU_Water_Cooling_Kit_New_Rev_4_Pump_Included_w_Free_Dead-Water.html?tl=g30c83s137

$500+ - Full Blown Custom loops

$1000+ Full Blown Custom Loops w/ rigid tubing, elbow fittings and lotsa gadgets and doodads

What did you do, what was ya reasoning and what did you get out of it ?

I am at the same stage as you. My conclusion so far is water cooling escorts the heat out of the case discretely where as CPU fans blow it thru the case and its up to the case fans to rid the heat but in the process heats up other components unnecessarily, also they add stress to the mobo.(cantilevered mass )

 

[JackNaylorPE]

Well not always ..... some radiators are mounted on intake fans ..... case bottom, case front .....see the H2O possibilities here.

http://phanteks.com/#&panel1-1&panel2-1

As to the stress, It's not a big deal I don't think unless you are trusting ya PC tot he FedEx gorillas, especially w/ MoBos w/ steel backplates (Asus Maximus VI Formula)

But theories aside, what do you expect to get out of ya WC ? What are your thoughts on single versus multiple loops ?
Thoughts on Radiators.... see links in 1st post ?

 

[JackNaylorPE]

Here's on way to think about it ....

http://www.youtube.com/watch?v=xt6SqfCLPiM 12:25

Looks like about $2k took him from 130.7 fps with stock intel cooler and stock clocks to 138 fps w/ WC and overclockingin BF3
Looks like about $2k took him from 27.5 fps with stock intel cooler and stock clocks to 36.3 fps w/ WC and overclockingin BF3

Cuda bought 2 Titans

 

[lazykoala]

My understanding is that WCing is not worth the time/money unless you plan on doing some extreme stuff with your build. It seems to be reserved for those who: A)have no budget limits, B) love the look of WC components, C) plan to max out most of their components. Today, components like GPU's are much improved and no longer sound like vacuum cleaners or jet engines, so air cooling can actually be quieter than WCed builds since most components aren't going to be run at 100% load anyway. If you run larger fans at lower RPMs or several smaller fans at low RPMs your PC can stay very very quiet. There is also the fact that most WC builds still depend on fans for airflow anyway and often have just as many fans as an air cooled only build.

 

[crooked windows]

Just how much better performance do we get moving from one of the better air coolers (Phanteks PH-TC-14, Thermalright Silver Arrorw, Noctua DH-14).

Been there done that!

Actually a custom kit in the 240 rad size category is on par with the big coolers you mentioned, which was a shock to me as I was expecting better than that after spending 3 times the money of a big air cooler.

That was back when I was attempting to cool a 2500K and get comfortably stable in the overclocked 5ghz range.

I found that water cooling for overclocking headroom had seriously changed from the days of the AMD FX57, and Ivy bridge is even hotter than the Sandy Bridge and now Intels new Haswell has taken the heat crown.

Traditional radiator loop water cooling has taken a kick in the groin when it comes to cooling these new CPUs and sometimes it amazes me to see those out here asking questions especially with serious overclocking intentions of their CPU and GPU and thinking it can all be done on one loop! :lol:

Not the MaMa!

No matter how many radiators are tied together in the cooling field the best you can reach is ambient temperature.

An overclocked CPU such as an Ivy Bridge gets hot enough as it is without adding the heat from the GPU to the loop, and IMO, delidding a brand new out of the box CPU for overclocking and instantly tossing the warranty away is sheer lunacy.

What did you do, what was ya reasoning and what did you get out of it ?

Edit: I forgot to say I think this is great you're doing this Jack, it will allow so much more shared information.

you mention "No matter how many radiators are tied together in the cooling field the best you can reach is ambient temperature"
what about wind chill factor? I honestly thought with a blowing wind, there would be a (in this case) a small drop below ambient.
you got me thinking now, and thats not good... my hair starts to fall out

 

[JackNaylorPE]

My understanding is that WCing is not worth the time/money unless you plan on doing some extreme stuff with your build. It seems to be reserved for those who: A)have no budget limits, B) love the look of WC components, C) plan to max out most of their components. Today, components like GPU's are much improved and no longer sound like vacuum cleaners or jet engines, so air cooling can actually be quieter than WCed builds since most components aren't going to be run at 100% load anyway. If you run larger fans at lower RPMs or several smaller fans at low RPMs your PC can stay very very quiet. There is also the fact that most WC builds still depend on fans for airflow anyway and often have just as many fans as an air cooled only build.

I don't think I can agree on the noise....but yes, CPUs have become more efficient and voltage limits are more likely to limit OCs than they were in the past. If you are pushing ya OCs to the max,noise definitely comes into play, especially on multiple GPU rigs with air cooling.

One of the attractions for me at least is that it opens the case interior to view and examination of all the components. I think it's akin to car enthusiasts where chrome exhausts / rims and engines ya can eat off of are a sense of pride and accomplishment....it doesn't get ya to work any earlier.

 

[crooked windows]

I have read a lot of opinions / youtube reviews too but few straight facts, and not all the facts in many cases leaving out as they are sponsored by them, so here is my pennies worth .
1 loop for the hot spot CPU to keep it neat and clean looking. Make the one loop do the best job by giving it the biggest /thickest rad possible. you mention adding a rad to the intake fan but isn't that putting extra heat into the case and against what we are trying to do. the lower the inside case air is, the faster / more efficient it will do its job. I plan on ducting the exhaust from the GPU card fans straight out the case with perspex or aircon tubing.
Multi loops are just adding to things that could go wrong (minuscule as it maybe). plus the savings can go to a better performing rad like an Aquacomputer or whatever is better
but if you are heading for "the look" of great cooling, go for it but each electrical gizmo added produces more heat and that to me is going in the wrong direction. this is of course dependent on your design layout.
More tubes = more air flow disruption (but that could work in your favor if designed well)
Go for biggest rad you can on top pulling heat out.

 

[JackNaylorPE]

I'll try and address one point at a time.....

At a certain point there is a point of diminishing or no significant return..... that's one of the purposes of these threads.... Going from 30 to 45 mm rad may show significant improvement but will 45 to 60 give the same....60 to 80 ?

Logic also sometimes conflicts with reality.....it was considered a given at one point in time that more water flow meant better cooling .... that larger tubes meant better cooling ..... that push was batter than pull in fans..... etc.... But actual testing showed that:

1. As long as ya meet minimum flow requirement ..... more gpm has no effect on cooling
2. Tube size has an insignificant effect on cooling, again....as long as minimum requirements are met
3. Push works better with hi speed fans, pull works better with low speed fans.

So, no I don't think a 4 x 140 mm x 80mm Monsta Rad with 8 fans will necessarily do a significantly better / quieter job than something smaller due to diminishing returns as ambient versus system temps difference diminishes. Most sites seem to recommend a 10C difference.

The aesthetic impact of a huge rad can also have be a negative . The Monsta rad in this build here for example looks "unbalanced" and cramped

http://themodzoo.com/forum/index.php?/page/articles.html/_/reviews/cases/phanteks-enthoo-primo-full-tower-review-r67?pg=6

I'm running closer to thermal limits on GPUs (84-85C) than I am on CPU limits on most systems so GPU cooling for me is a high priority. Im also finding that blowing all that air thru a single grilled slot opening via hi speed 60mm fans in noisy and inefficient. A half of full dozen large low rpm case fans can do this more quietly and effectively.

Another way to look at it.... is it worth say investing in $500 cooling loop for GPUs ? Answer can vary. If ya planning two 770s, then Id say that $500 would be better spent with putting the $500 into two 780s........ If ya already planning 780's the answer is bit different as Titans are $700 addition .... if ya got Titans, then ... well its ya only option.

As to electrical gizmos adding more heat....well we could let convection cool our rads but I think we'll all agree that the heat balance favors using fans even tho they produce a lil heat themselves.

Even aesthetics comes into play. I love the look of straight tubing and right angles ..... and doing a 90 degree bend w/ Bitspower fittings is $40 for the 3 pieces.,,,,and that's a threshold I don't wanna pass unless absolutely necessary.

What Im am trying to explore here is thresholds .... the questions being asked much depend on where ya sitting, what ya goals are and what ya options are. There are no universal answers. So yes, ya can do WC for performance goals, for noise reductions, for aesthetics or simply because it's a "hobby" you enjoy..... what "makes sense" will largely .... no wholly .... depend on where ya looking at it from.

 

[JackNaylorPE]

If you want to mist water over your radiators while they are running you could create a wind chill effect, but it would be rather messy, with electrical shorting possibilities, and possibly a serious mildew producer.

If your air is 72 degrees and the wind chill is -10 below that, it may "feel like" 62 degrees but a bottle of soda left on the picnic table outta the sun will never get below 72

The effect you describe represents the phase change from evaporation which is also what makes air conditioners work....changing from liquid to a gas absorbs heat.... but that's not wind chill.

 

[JackNaylorPE]

Ok, it's not wind chill effect ..... It's evaporative cooling effect

http://www.ehow.com/how-does_5315235_evaporation-cause-cooling.html

When a liquid evaporates, its molecules convert from the liquid phase to the vapor phase and escape from the surface. What drives this process is heat. In order for the molecule to leave the liquid surface and escape as a vapor, it must take heat energy with it. The heat that it takes with it comes from the surface from which it evaporated. Since the molecule is taking heat with it as it's leaving, this has a cooling effect on the surface left behind.

http://science.howstuffworks.com/nature/climate-weather/atmospheric/question70.htm

For an inanimate object, windchill has an effect if the object is warm. For example, say that you fill two glasses with the same amount of 100-degree water. You put one glass in your refrigerator, which is at 35 degrees, and one outside, where it is 35 degrees and the wind is blowing at 25 mph (so the windchill makes it feel like 8 degrees). The glass outside will get cold quicker than the glass in the refrigerator because of the wind. However, the glass outside will not get colder than 35 degrees -- the air is 35 degrees whether it is moving or not. That is why the thermometer reads 35 degrees even though it feels like 8 degrees.[/b]

 

[JackNaylorPE]

Here's a perfect example of "does it make sense ?"

http://www.anandtech.com/show/7363/the-neophytes-custom-liquid-cooling-guide-how-to-why-to-what-to-expect

After about $800 of WC equipment.......

http://www.anandtech.com/show/7363/the-neophytes-custom-liquid-cooling-guide-how-to-why-to-what-to-expect


Moving from Air to Water and CPU Temp went up 5.1C !

At least GPU Temp went down 4C

I'm not saying the author / builder made great decisions, but in the Haswell era, if ya thinking of huge drops in CPU temps, and significant performance increases, ya need to have realistic expectations.

 

[4Ryan6]

Here's a perfect example of "does it make sense ?"

http://www.anandtech.com/show/7363/the-neophytes-custom-liquid-cooling-guide-how-to-why-to-what-to-expect

After about $800 of WC equipment.......

I'm not saying the author / builder made great decisions, but in the Haswell era, if ya thinking of huge drops in CPU temps, and significant performance increases, ya need to have realistic expectations.

That's a very good example of what I've been trying to say regarding the 2nd, 3rd, and 4th, generations of Intel CPUs, and tying full coverage GPU cooling blocks all in a single loop, it's a bad idea!

Plus he added the motherboards heat to the loop as well, so the outcome is falling right in line with what it should be, he actually needed three times the radiator cooling field than he actually used, but what charts and calculators indicate that exactly? None!

An air cooled graphics card handles the GPU heat load and the VRs and Memory heat load on separate heat sinks, all cooled with air flowing over them, when that is replaced by a full coverage water block, the GPU, VRs, and Memory is all cooled by the same water block, the coolant transfers the heat from it all, to the radiators.

Add a 2nd full coverage graphics card and you have some serious heat dissipation to get rid of, add that to a hot to overclock CPU and you're simply pissing into the wind, for a simply sick reference!

That kind of setup needs to be a dual loop setup any way you look at it, IF, the builder is after overclockable CPU cooling performance, because that's just a for looks setup, not a performance setup, IMO.

Secondly the guy can assemble a cooling loop obviously, (I'm surprised he didn't attempt to cool his car at the same time!) but he is clueless of what it actually takes to properly cool his overclocked CPU to gain overclocking headroom, and rise above the stock GPU air cooling.

The CPU should be stand alone cooled with zero heat producers added to the loop and for some descent overclockable results on a minimum of a 120 quad rad.

My suggestion, from My own discoveries, overclocking these type CPUs!

Go less you get less, anyway you look at it, you cannot get something for nothing, there is always consequences!

Ryan

 

[manofchalk]

Yeah, looking at that article he has way insufficient rad space. Overclocked 4770k, dual 780's and VRM heat with 360+240mm of slim rad space, anyone who knew what they were doing could have predicted this outcome.
That, and he has all of this in series with a single DDC pump. You cant exactly say he made his loop that optimized for performance here.

But yes, it does show that water on graphics cards has a far greater benefit than on the CPU. We really need to get in the habit of recommending people water-cool their graphics cards, and then include their CPU later on rather than the other way around.

 

[4Ryan6]

Yeah, looking at that article he has way insufficient rad space. Overclocked 4770k, dual 780's and VRM heat with 360+240mm of slim rad space, anyone who knew what they were doing could have predicted this outcome.
That, and he has all of this in series with a single DDC pump. You cant exactly say he made his loop that optimized for performance here.

But yes, it does show that water on graphics cards has a far greater benefit than on the CPU. We really need to get in the habit of recommending people water-cool their graphics cards, and then include their CPU later on rather than the other way around.

I agree but it seems lately too many are asking for advice and are clueless as to what they'll even be cooling or what their intentions are for considering water cooling in the first place.

Many see some of the high CPU overclocks out there and think they can get the same, but do not understand how they got those high clocks, or how it was cooled to do it.

Then they develop the mental attitude of, well?, if they got that far surely with good water cooling I can get at least to this level?, not realizing the high overclock was done using LN2, Phase Change, Dry Ice, Direct Peltier Mount, or a Chilled water system.

Because many boasting high CPU overclocks, don't divulge how they got there.

How many times do we see disappointment because their vast water cooling investment does not cool as well as their high quality air cooler did, simply because it was all crammed into one loop.

That reality is a hard pill to swallow after a serious amount of money has been spent on a water cooling solution!

Been there done that!

That's why I try to keep them from making the same mistakes.

 

[JackNaylorPE]

But there's one huge point left out...... heat is clearly not the performance limiter on modern CPUs. I'm not talking how high ya can get long enough to get validation before ya system crashes..... I'm talking 24/7/365. My son's 2600k w/ twin GFX in SLI has been running on air at 4.8 since a month after SB came out ..... It runs just fine at 5.0 but at a voltage I wasn't comfortable with for long term use. Just how far can ya get incrementally on water.....throwing more and more rads at it.

I think us geeky types spend to much time focusing on the temps and not enough on the performance results. He dropped 40C off his GPU temps.... but where did it get him .... 1 or 2 fps on the only games that were under 60 fps ? Was that worth $800 ? If he doubled his investment, would he show more ?

On the synthetic CPU oriented benchmarks...... even tho his temps went up, so did his performance.....but with Ivy Bridge and Haswell the heat curve is so danged steep that the scale of diminishing returns is extreme. I'd love to see some plots of voltage versus Ghz on several HW samples and then compare them to IB. I mean we've all read the articles, I just don't have a good sense of how steep that curve is cause results are all over the place.

 

[manofchalk]

I agree, water-cooling is not objectively worth it.
But we still get it anyway, because its pretty damn cool.

 

[4Ryan6]

But there's one huge point left out...... heat is clearly not the performance limiter on modern CPUs. I'm not talking how high ya can get long enough to get validation before ya system crashes..... I'm talking 24/7/365. My son's 2600k w/ twin GFX in SLI has been running on air at 4.8 since a month after SB came out ..... It runs just fine at 5.0 but at a voltage I wasn't comfortable with for long term use. Just how far can ya get incrementally on water.....throwing more and more rads at it.

I think us geeky types spend to much time focusing on the temps and not enough on the performance results. He dropped 40C off his GPU temps.... but where did it get him .... 1 or 2 fps on the only games that were under 60 fps ? Was that worth $800 ? If he doubled his investment, would he show more ?

On the synthetic CPU oriented benchmarks...... even tho his temps went up, so did his performance.....but with Ivy Bridge and Haswell the heat curve is so danged steep that the scale of diminishing returns is extreme. I'd love to see some plots of voltage versus Ghz on several HW samples and then compare them to IB. I mean we've all read the articles, I just don't have a good sense of how steep that curve is cause results are all over the place.

I agree, water-cooling is not objectively worth it.
But we still get it anyway, because its pretty damn cool.

I agree completely with both your statements as to relating to traditional radiator water cooling!

For the record however the cooling I am using goes way beyond the cooling capabilities of what an ambient cooling solution does, and knowing what it can do I cannot agree with those statements.

I can run 24/7 stable temperatures far greater than radiator cooling can touch so being able to keep it cool enough does make a major difference.

A 3770K @ 5ghz that has not been delidded!

Cooling made that possible, but not radiator cooling!

 

[manofchalk]

True, but you are outside the norm here Ryan :lol:
Your going to have to come up with a name for what your doing, Peltier Cooling would be the easy one but has the history of having Peltiers being stuck directly to chips, which is nowhere near what your doing.
Water-TECooling?

 

[4Ryan6]

True, but you are outside the norm here Ryan :lol:
Your going to have to come up with a name for what your doing, Peltier Cooling would be the easy one but has the history of having Peltiers being stuck directly to chips, which is nowhere near what your doing.
Water-TECooling?

You're right! MOC, It's way outside the norm!

For something that started as an experiment to be as successful as it is I want to shout it from the rooftops, but it is just completely beyond 99% of the water cooling community.

I guess it does need it's own name as it is new territory and it really has no place in this thread, other than mentioning it's cooling capabilities.

I looked back to when this cooling was successfully put in operation and I've actually been running this cooling now for 1 year and 3 months.

I've been determined to run below ambient but above zero staying just above the condensation forming point, which where I live in South Carolina is a dependable 13c below ambient.

There's cooling capabilities at those temperatures that has been basically a learning adventure all it's own, especially regarding 24/7 overclocking.

Water-TECooling, That's not bad!

 

[4Ryan6]

Been a lot of questions on the board of late on water cooling and having spent lotta time looking at stickies around the web, I'm noticing lotta information outdated and some still not there.

You are 100% correct, at least from what I've found so far.

The new CPUs especially relating to Intel's 3rd, and 4th, generation has been the hot topic of overclocking with more details regarding delidding CPUs to reach higher clocks.

To my knowledge overclock vs heat without delidding relating to radiator water cooling information is MIA.

That's why some after spending a chunk of money on water cooling end up disappointed, as they're relying on old information.

The past recommended radiator size or cooling field, is just not going to do the job with these hot CPUs, if overclocking headroom is in the picture, go big to what the majority call overkill, or go home and cry, because the end results fall way short of the assumed goal.

Too many are still recommending a 240 even a single pass to overclock an Ivy Bridge or Haswell, That's a joke! :lol:

Many because they're determined to cram it all in one case!

IMO the minimum size radiator for overclocking those CPUs is a 480 at least a dual pass, tight fin requiring comparable high static fans, and that's for the CPU alone, and that's a MINIMUM recommendation.

If the cooling field is not there to handle the load, neither is the cooling!

 

[4Ryan6]

But there's one huge point left out...... heat is clearly not the performance limiter on modern CPUs.

Totally disagree! (when below ambient cooling is used)

Heat is the performance limiter!

Why do you think they delid their CPUs, to lower the heat!

Below is a quote from the Exploring Below Ambient Thread.

Without this cooling I could not run my 2500K at 5ghz for anything more than a CPU-Z validation, but this cooling allows not only reaching but acquiring stability once getting there. The thermal mass of the water ensures longevity of certain temperature levels that you may be needing to accomplish your overclocking goals successfully. Additionally for longevity the system is constantly running at cooler levels all the time, even if I allowed the water temperature to rise to ambient room temperature, it still bests a standard water cooling closed loop, it opens completely new doors of possibility.

Especially to a 24/7 stable overclock of the new CPUs!

Regarding traditional radiator cooling, you have to have a minimum of 50% more cooling field than needed to do the job!

Even then you are still at the mercy of ambient.

Water Cooling 101 - Should I or Why Bother ?

The title to this thread makes it about more than just radiator cooling, there are other options out there that will give excellent new CPU 24/7 stable overclocking results.

May not be the cooling options for everyone, but they're still out there, and being used everyday.

I've been exploring these cooling alternatives since June of 2011.

The CPU-Z validation to the right is my 24/7 overclock of my 3770K (Not Delidded)

 

[JackNaylorPE]

1. I'm talking " modern" as in Haswell .... Not last generation IB ..... different world. The silicon lottery has far greater impact on HW than it did on IB. It's very, very possible to have one HW chip hit 4.8 on Air and another hit a wall at 4.5 with delta 10C water.

2. The voltage versus GHz curve is way way steeper on Haswell than it was on IB. To get a 0.1 Ghz improvement on Haswell we are seeing much larger voltage increases required. SB was cake to get tob 4.8 Ghz.... if ya particular CPU was capable, 5 GHz was no issue in air.

3. Below ambient isn't really isn't in the context of generic "water cooling". At least that wasn't the intent. Intent was "water only" not water + [insert other technologies here]". Within the context of the thread, I was looking to address basically "set it and forget it" type of ownership for PC's that would be used in an office or home environment. I very much enjoyed your thread on below ambient but that's not a box i would use outside of a workshop type setting.

BTW, I regarding a post in ya below ambient thread by someone:

Good air cooling will allow 4.5GHz on the 2500k but not 5.0GHz

That surprised me..... been there, done that. Not unusual.

http://archive.benchmarkreviews.com/index.php?option=com_content&task=view&id=674&Itemid=62&limit=1&limitstart=5

It's almost ironic that coolers like this are becoming available just as processors transition to designs that may ultimately render them unnecessary; even overclocked to 5GHz, an Intel Sandy Bridge 2600K doesn't need anywhere near this level of cooling. Still, it wouldn't hurt, and as I noted earlier, there are still CPUs out there that can benefit from it.

Been sort of a roller coaster for cooling over the years..... when WC rose to be talked about in the "mainstream", Intel was fighting OC'ers, then later embraced them peaking with SB where 50% OC's were common. As above, I started to wonder if WC would fade away w/ 5 GHz on air but IB gave it a shot in the arm However Haswell's huge unit to unit variation and extraordinarily steep voltage curve puts a big damper on it because they're hitting a wall regardless, even if heat isn't yet a serious concern.

With the 2600k .... these were 1st attempt numbers....tweaking later brought them down.

-I needed a 0.024 voltage increase to go from 4.2 to 4.4 while temps went up 2-3C
-I needed a 0.080 voltage increase to go from 4.4 to 4.6 while temps went up 6-8C
I needed a 0.048 voltage increase to go from 4.6 to 4.8 while temps went up 7-11C

But the point I am making is that w/ Haswell we need much bigger voltage increases and therefore we quickly get to unhealthy voltage levels that many are not comfortable running at 24/7/365. On a machine that needs to "be there" cause ya make a living with it or it's damn inconvenient to have ya only entertainment and communications outlet not available, ya don't want to send ya CPU to an early demise from over volting which as you know has other issues besides heat.

On these and other boards see loads of posts where peeps hit a wall at 4.4 or 4.5 and their temps are just fine. I have not personally had an opportunity to push the limits Haswell as yet as I only get to do minor OS's on boxes that are out the door quickly....that will change shortly as have major build in planning stages.

I'm hoping we start seeing better consistency between CPUs as the production line matures.

JJ reports in this video that they have seen 4.6 Ghz at anywhere from 1.15 to 1.40 volts. That 1st one will do 4.8 on air .....the 2nd one, I don't think it will do 4.8 on water no matter how many Rads ya throw at it. Id be scared to run it at 4.6.

http://www.youtube.com/watch?v=Z7zPu9255ZI @ 13:30

 

[4Ryan6]

1. I'm talking " modern" as in Haswell .... Not last generation IB ..... different world. The silicon lottery has far greater impact on HW than it did on IB. It's very, very possible to have one HW chip hit 4.8 on Air and another hit a wall at 4.5 with delta 10C water.

2. The voltage versus GHz curve is way way steeper on Haswell than it was on IB. To get a 0.1 Ghz improvement on Haswell we are seeing much larger voltage increases required. SB was cake to get tob 4.8 Ghz.... if ya particular CPU was capable, 5 GHz was no issue in air.

My bad!, Didn't see the thread only covered Haswell?

Below ambient isn't really isn't in the context of generic "water cooling".At least that wasn't the intent. Intent was "water only" not water + [insert other technologies here]".

Just adding my 2 cents Jack, didn't see generic "water cooling" in the thread title or first post to indicate this subject being specific.

I have not personally had an opportunity to push the limits Haswell

That explains some of your response, when you get some actual experience in it, lets talk!

Pretty much the same assumptions that tagged Ivy Bridge.

There are other Haswell and Ivy Bridge overclock limiters besides temperature. like improper BIOS settings, overclocking their CPU memory controllers, and generally a lack of overclocking knowledge, so whether someone can get higher multipliers or not may be their own fault that proper air or water cooling could handle.

On a machine that needs to "be there" cause ya make a living with it or it's damn inconvenient to have ya only entertainment and communications outlet not available, ya don't want to send ya CPU to an early demise from over volting which as you know has other issues besides heat.

Jack I cannot speak for others but for myself, I do not overclock any of my critical aspect machines, so there's no need to water cool them!

I only overclock what I directly take responsibility for, and that is my gaming machine period.

Additionally, my Peltier/TEC cooling is set it and forget it, but I would not use that on a critical aspect computer or work computer, because I would not be overclocking such a critical machine!

Below ambient cooling can become a cheaper and very effective cooling option, it does not have to be as expensive as my present setup, it's been done with a cheap styrofoam cooler, CPU water block, and an aquarium pump, and Ice. It can be done with less money invested than some of the CLC water coolers, so it is another water cooling option.

Water Cooling 101 - Should I or Why Bother?

Shouldn't that cover all aspects of water cooling?

If it can be done, what's wrong with making others aware of it?

I've been blasted with negativity since I first opened the Exploring Below Ambient Water Cooling thread, so it's not like I'm not used to it!

There are cooling options besides traditional radiator water cooling, and whether someone decides to use those other possible cooling options, should be their own choice to make.

Below ambient cooling breaks the mold of what's been established and accepted in the water cooling realm, but it breaks the barrier of confinement and makes traditional statements untrue, if you regard water cooling as not confined to radiator water cooling.

The quote below was taken from Overclockers.com, from the Guide to Delta-T in Water Cooling, posted on 6/4/2010, by Conumdrum.

Excellent information and the statements are true relating to traditional radiator water cooling, but not so when a below ambient cooling solution is used because ambient temperature and radiator output temperature is no longer a part of the equation.

Delta T (DT) and Why it’s so Important to Understand it

DT is the foundation of your water cooling loop. The better your DT, the cooler your chips are. In water cooling, DT is simply the difference between the ambient air temperature and the water temperature on the outgoing side of the radiator. Room temperature vs. water temperature: that’s it. You can’t remove all the heat – no system is 100% efficient, nor can you go below ambient room temperature.

When you boot up a powered off, water-cooled PC, the water and your CPU are at room temperature. When you boot the PC up, the chip gets hot very fast. The water moves over the chip, it begins to remove heat, the heat goes to the radiator, and some of the heat is removed. Not all of it can be removed. You have to know a lot of thermodynamics theory deeply (more than me) to know exactly why. The water begins to warm up slowly, and in time it reaches a balance: an equilibrium. Heat is made and heat removed, the loop is stabilized and temperatures will not change.

If you change the room temperature, the load on the loop or your fan speed, the loop needs to readjust. This is when we like to measure our cooling ability – usually 30 minutes at a stable load is long enough to begin to measure. If you increase your cooling capability, the water will get cooler.

Water temperatures in a stabilized loop, amazingly, are very similar anywhere in the loop. There is only a 2-3°C maximum difference between the radiator out temperature and the CPU out temperature; this has been verified by Skinnee. Remember, the water can’t remove all the heat, some is transferred to the air. Your radiator size, efficiency and fans play a big part in this. Look at it this way – it’s a system built on many parts and within the laws of physics. Every part affects the other.

The underlined in his statement is true with radiator cooling, but not with below ambient cooling, to Conumdrum's credit experimenting with below ambient water cooling was not necessary with the CPU examples referenced in the rest of his article.

Specifically the 775 platform, as I was able to overclock my Q9550 to 4ghz on air cooling, and AthurH got his to 4.25ghz, using ice chilled water, but that's another story.

Below ambient cooling simply breaks the rules, and allows much more than traditional radiator cooling ever could!

Isn't that what you meant in your opening post when you said?

as goals here are WC systems that provide better performance and lower noise than air coolers