Water cooling vs. Fan cooling.

[StefanR5R]

I would have to switch off and move the box for this. This is not possible today, but maybe tomorrow or some other day.

 

[bigboxes]

I would have to switch off and move the box for this. This is not possible today, but maybe tomorrow or some other day.

Cool. I wish more would share pics of their rigs and the topic they are talking about. Thanks!

 

[BonzaiDuck]

I am quite new to water cooling myself. I have a 240mm AIO running my 3570k @4.5ghz and a 280mm AIO still in the box for my next build. In the past I have always used 120mm air coolers. I still have the thermalright Ultra 120 and recently bought ($12) an upgrade kit for it to use on 115x motherboards. I have two ultra 120's circa 2007/08, one in use on a G4560 (max temps in low 40's).

The Coolermaster 212+ is based on the original Thermalright Ultra 120 and most of the current 120mm air coolers based on that design as well. They work well for OCing CPU's.

In my opinion custom loops are cool but not cost effective. Something a person who wants to have a custom loop builds for the sake of saying they have it or they did it. Not the most reliable parts for cooling solutions.

I personally think any 240mm AIO cooler is superior to any of the Air coolers. I am not a big fan of Noctura air cooling simply for the cost of the cooling solution. Based on my extensive experience with air cooling, I have found AIO liquid cooling to be superior. My 3570K @4.5ghz has never touched 60C on any core. With air cooling my temps on a dust free Xigmatek Gaia 120mm air cooler were mid 60's to high 60's. All with Arctic Silver 5 thermal paste. My AIO 240mm liquid cooler (3570K @4.5ghz) typically has temps at 50C or slightly below with a few cores in the low 50's. during heavy gameplay of BF1. I use Core Temp for my readings and the highest a core has hit on my 3570K was 59C. Typically the hot core is 53-56C in heavy gaming.

One of the arguments about water cooling is that the best air coolers can catch up to AIO air coolers after marathon gaming sessions 4 hours or more. I suspect that could be the case with 120mm air coolers but the 240mm air cooler seems to efficiently pump out the warm liquid to the radiator. In my liquid cooler it seems whenever the temps appear to get high mid 50's the liquid cooler pumps out the warm liquid filtering in much cooler liquid from the radiator.

If a person doesn't OC, there is no reason for liquid cooling solutions unless you are doing it for the cool factor. The useful life of a GPU is shorter than the life of a CPU. After a GPU is retired from a gaming system, you would have to reinstall the air cooling heatsink and fans. Problems typically occur in this process. Again all based on the skill of the computer builder. Taking off the good air cooling of GPU's could result in damage to the GPU, it happens. Typically those who do liquid cooling with CPU and GPU do it because they have a custom loop in their system.

When I installed my 240mm AIO cooler. I crafted a back plate for my GTX 970 to catch any leaks that may occur.

I would come to approximately the same conclusions looking at it from the opposite side. I had planned to build my Skylake system with an EKWB Predator or Swiftech H240 X2 customizable or expandable AIO. Then I saw some review comparisons including the Predator, the NH-D15 and the ThermalRight Le Grand Macho -- hereafter referred to as LGM. The LGM bests the D15 by 1 or 2C; the EKWB bests the LGM by 5C.

Delidding the processor and then restoring it with CLU (Liquid Ultra) for the TIM, I get an immediate 12C advantage. So, compared to the EKWB using an unmolested retail-box CPU, I come away about 7C better off. If I can't reach 4.8 Ghz on my Skylake, it is not because of the cooling solution and CLU-relidding, but because of a self-imposed voltage limit or voltage target which I do not want to exceed. In fact, I CAN "get to 4.8" by barely exceeding my own limit by 10 mV beyond my limit and it will pass several different stability tests. It just may not pass all of them unless I'm I add another 10 mV after that. So I'm fine with what I've got.

You CAN get some competitive overclocks with air-cooling and maybe a "cheat" for replacing the processor's TIM. If you plan on building a hexa- or octo-core system (i.e., X99 chipset LGA-2011v3), you can't relid with CLU because the processor was fabricated with indium solder instead of a gummy polymer. So you're well-advised to consider water-cooling if you plan to seriously overclock.

Now -- I could benefit further by capturing another 5C to 8C degrees advantage by replacing my LGM with a Predator, H240 X2 Prestige, NZXT Kraken x62, or any AiO in their class. But the only way it would be worth it to me requires that I decide to exceed my voltage prohibition. So I'm unable to decide, and no less satisfied with what I have.

 

[StefanR5R]

There is really no surprise that many AIOs can just so keep up with big air coolers. Just compare the total fin surface of typical AIOs and big air coolers. Result: The latter have the same amount of fin surface, in some cases even more.

The difference between AIOs and air coolers is that AIOs use flexible water filled tubes, not rigid heatpipes, to connect the cooling block with the radiator. Ideally, this is used to place the AIO's radiator into an aerodynamically optimal location.

The difference between custom water loops on one hand, and AIOs and air coolers on the other hand, is that humongous amounts of radiator surface can be added. Furthermore, there is more flexibility WRT which components to put a cooling block on. And it is possible to place the radiator outside of the computer case.

Also, let me stress again, watercooling a sub-120 W CPU does not make a lot of sense. But watercooling a 100 W+ GPU makes much sense, simply due to the space constraints around GPUs. Everytime when the question arises whether or not to watercool, the GPU should always be thought of first. The CPU should be an afterthought at most, unless we are talking about overclocked 8-core CPUs for example.

IMO.

 

[Hans Gruber]

The LGM heatsinks looks very much like the old Tuniq Tower 120 but with the fan strapped on the outside of the heatsink. I am with you on the benefits of delidding your CPU with liquid metal. I know on the Kaby Lake processors they see temp improvements up to 20C. I think the max OC within reason for a Skylake 6700K is right around where you are @ 4.8ghz

My opinion on the AIO water cooling solutions is that 240mm radiators is the minimum for OCing and anything less 120mm is really just for show as in the cool factor to have. Air Cooling can beat or match any 120mm liquid cooling solution. The other benefit to liquid cooling is that you may be able to lower your voltage and achieve the same OC as your LGM air cooler.

With that said, I took out of retirement a Thermalright Ultra 120 that was cooling a Q6600 @ between 3.3ghz and 3.6ghz way back in the late 2000's. I have another Thermalright Ultra 120 that I will put an adapter kit on for future use on other builds. The current Ultra 120 is mated with a G4560 budget build that I just finished. The upgrade mounting kit was $12.

I do see your point with regards to your OCing. My 3570k runs @4.5ghz without any speed step or any power saving features because it's a desktop gaming rig. I roll my eyes when I watch youtube videos of builders showing off their OC'd system and the CPU throttles down from whatever they have it OC'd at to half the speed for a few seconds.

 

[MrTeal]

There is really no surprise that many AIOs can just so keep up with big air coolers. Just compare the total fin surface of typical AIOs and big air coolers. Result: The latter have the same amount of fin surface, in some cases even more.

The difference between AIOs and air coolers is that AIOs use flexible water filled tubes, not rigid heatpipes, to connect the cooling block with the radiator. Ideally, this is used to place the AIO's radiator into an aerodynamically optimal location.

The difference between custom water loops on one hand, and AIOs and air coolers on the other hand, is that humongous amounts of radiator surface can be added. Furthermore, there is more flexibility WRT which components to put a cooling block on. And it is possible to place the radiator outside of the computer case.

Also, let me stress again, watercooling a sub-120 W CPU does not make a lot of sense. But watercooling a 100 W+ GPU makes much sense, simply due to the space constraints around GPUs. Everytime when the question arises whether or not to watercool, the GPU should always be thought of first. The CPU should be an afterthought at most, unless we are talking about overclocked 8-core CPUs for example.

IMO.

I'd agree with your opinion. The ATX standard is designed such that there is a massive amount of room around the CPU for a big tower cooler. You'll get some improvement with a AIO or custom water, but the benefit can be pretty marginal compared a good air cooler and the noise profile of a lot of AOIs is worse than a top air cooler.

GPUs on the other hand are inherently limited by the 1.5" spacing you get between expansion slots for a double slot card. Not only don't you have much volume for fins, but the long restricted channel means the air is going to have to be forced around some tight bends. Granted my setup is larger than most, but if someone gave me $100 to spend on cooling for a system, I'd probably stick with the stock CPU cooler and buy a AOI and bracket combo for the GPU.

 

[aigomorla]

sigh...

ok first off AIO != true watercooling.

A true water cooling is a system which uses a pump that will generate a flow greater then .75gpm.
Why? because if one does the math to thermo properly, the carrying potential of water gets closer to the apex of its heat curve as you approach .75gpm and pretty much flattens out as you reach 2gpm.

So that means the more flow one has in a LCS system past .75, the better optimized water gets until you hit about 2gpm, where diminished returns starts kicking you in the butt.

AIO's at best generate .25-.5gpm which puts there efficiency at the rock bottom.
At best one should only consider a AIO if you have height constants on a 6-8 heat pipe air cooler, as those will probably be a lot safer + make less noise, and have less potential of crying when something breaks and leaks all over your expensive hardware.


That being said, if were talking about an intel processor which is not hard soldered with IHS, watercooling basically brings almost no merits into our hobby. The resistance of the TIM in between the IHS and DIE on a LGA115x processor is so horrible, that watercooling will not give you an edge. So unless u want to Delid that cpu, i would not recommend a LCS system on that cpu, especially since the mosfets around the CPU require airflow, and that is again something which watercooling wont generate.

If were looking at LGA2011, thats a different story. With a good system, there is no way an aircooler or AIO can compete.
1.5gpm (the gold zone) has a carrying potential of 300W / degree C. That means it takes 300W of heat to make the water go up 1C.
Pair that up with a 120x3 radiator with 1800RRM, your system will bascially have equalibirum of 2-3C over ambient in a heat load of less then 300W.

But watercooling i feel has shifted more to GPU then CPU.
You net a heat reduction of load temps by 1/3-1/2.... that means those 65C dies will at best only be about 35-40C.
Full Cover blocks also cool RAM + VRMS, to extend the life of the card. It was stated that every 10C you lower a temp in an IC, you not only make the circuit more efficient, but you also double the life of the chip. But again, unless ur cooling a 1080, or a expensive top tier GPU, a full cover block is not worth it over an upgrade to the next platform..
So keeping something cooler, can in sense, net you a better clock in overclocking.

1070GTX (for example) should not be watercooled, as it is better to invest the $$$ of the full cover block to get a 1080GTX and air cool it.

 

[BonzaiDuck]

sigh...

ok first off AIO != true watercooling.

A true water cooling is a system which uses a pump that will generate a flow greater then .75gpm.
Why? because if one does the math to thermo properly, the carrying potential of water gets closer to the apex of its heat curve as you approach .75gpm and pretty much flattens out as you reach 2gpm.

So that means the more flow one has in a LCS system past .75, the better optimized water gets until you hit about 2gpm, where diminished returns starts kicking you in the butt.

AIO's at best generate .25-.5gpm which puts there efficiency at the rock bottom.
At best one should only consider a AIO if you have height constants on a 6-8 heat pipe air cooler, as those will probably be a lot safer + make less noise, and have less potential of crying when something breaks and leaks all over your expensive hardware.


That being said, if were talking about an intel processor which is not hard soldered with IHS, watercooling basically brings almost no merits into our hobby. The resistance of the TIM in between the IHS and DIE on a LGA115x processor is so horrible, that watercooling will not give you an edge. So unless u want to Delid that cpu, i would not recommend a LCS system on that cpu, especially since the mosfets around the CPU require airflow, and that is again something which watercooling wont generate.

If were looking at LGA2011, thats a different story. With a good system, there is no way an aircooler or AIO can compete.
1.5gpm (the gold zone) has a carrying potential of 300W / degree C. That means it takes 300W of heat to make the water go up 1C.
Pair that up with a 120x3 radiator with 1800RRM, your system will bascially have equalibirum of 2-3C over ambient in a heat load of less then 300W.

But watercooling i feel has shifted more to GPU then CPU.
You net a heat reduction of load temps by 1/3-1/2.... that means those 65C dies will at best only be about 35-40C.
Full Cover blocks also cool RAM + VRMS, to extend the life of the card. It was stated that every 10C you lower a temp in an IC, you not only make the circuit more efficient, but you also double the life of the chip. But again, unless ur cooling a 1080, or a expensive top tier GPU, a full cover block is not worth it over an upgrade to the next platform..
So keeping something cooler, can in sense, net you a better clock in overclocking.

1070GTX (for example) should not be watercooled, as it is better to invest the $$$ of the full cover block to get a 1080GTX and air cool it.

And that all makes sense. It then boils down to your choice as to a balance between what you "need" in a system and what you "want." I'd had two or three "plans" on the table for "E" processor hexa/octo-core systems variously with X79 or X99 chipsets. But in practical terms, I can't use all of that processing power much of the time, so those plans went into PDF archive and haven't come out. Those times when I could really use it I can probably get by with an i7 or i5 K processor.

The other problem in an area I don't incline toward but probably practice anyway, is the size of the case. A lot of folks are building small-form-factor machines. I just choose not to build any more with full-tower cases. In my midtowers, I can manage a radiator of 285 to 300mm in length.

I think we also had a discussion more than a year ago about radiators like the MO-RAx models. Similar designs could be found with other water-cooling parts makers. With that, you lose something of computer mobility but for quick-release fittings, and you still have to move two items around. You need between 8 and 18 or 32 fans to have push-pull with those units-- even with 180mm fans, so you have to make up the cabling and refinements.

That only leaves the question as to whether the future processors with six or eight cores will come with lower TDP with indium solder. Lowering TDP while reducing the die size and lithography has mixed results. And it is amazing to see how much cooler a chip runs with CLU under the lid as opposed to Intel's polymer goop.

 

[StefanR5R]

Re #25, no photo downtime today, and apparently not before May 19.

 

[BonzaiDuck]

I am quite new to water cooling myself. I have a 240mm AIO running my 3570k @4.5ghz and a 280mm AIO still in the box for my next build. In the past I have always used 120mm air coolers. I still have the thermalright Ultra 120 and recently bought ($12) an upgrade kit for it to use on 115x motherboards. I have two ultra 120's circa 2007/08, one in use on a G4560 (max temps in low 40's).

The Coolermaster 212+ is based on the original Thermalright Ultra 120 and most of the current 120mm air coolers based on that design as well. They work well for OCing CPU's.

In my opinion custom loops are cool but not cost effective. Something a person who wants to have a custom loop builds for the sake of saying they have it or they did it. Not the most reliable parts for cooling solutions.

I personally think any 240mm AIO cooler is superior to any of the Air coolers. I am not a big fan of Noctura air cooling simply for the cost of the cooling solution. Based on my extensive experience with air cooling, I have found AIO liquid cooling to be superior. My 3570K @4.5ghz has never touched 60C on any core. With air cooling my temps on a dust free Xigmatek Gaia 120mm air cooler were mid 60's to high 60's. All with Arctic Silver 5 thermal paste. My AIO 240mm liquid cooler (3570K @4.5ghz) typically has temps at 50C or slightly below with a few cores in the low 50's. during heavy gameplay of BF1. I use Core Temp for my readings and the highest a core has hit on my 3570K was 59C. Typically the hot core is 53-56C in heavy gaming.

One of the arguments about water cooling is that the best air coolers can catch up to AIO air coolers after marathon gaming sessions 4 hours or more. I suspect that could be the case with 120mm air coolers but the 240mm air cooler seems to efficiently pump out the warm liquid to the radiator. In my liquid cooler it seems whenever the temps appear to get high mid 50's the liquid cooler pumps out the warm liquid filtering in much cooler liquid from the radiator.

If a person doesn't OC, there is no reason for liquid cooling solutions unless you are doing it for the cool factor. The useful life of a GPU is shorter than the life of a CPU. After a GPU is retired from a gaming system, you would have to reinstall the air cooling heatsink and fans. Problems typically occur in this process. Again all based on the skill of the computer builder. Taking off the good air cooling of GPU's could result in damage to the GPU, it happens. Typically those who do liquid cooling with CPU and GPU do it because they have a custom loop in their system.

When I installed my 240mm AIO cooler. I crafted a back plate for my GTX 970 to catch any leaks that may occur.

something like a baking pan or cookie-sheet with a raised edge?

 

[BonzaiDuck]

Here's another thought. Aigomorla notes that you might double the life of your CPU for every 10C reduction in temperatures.

At stock settings, with the CPU's own TCASE spec for OEM builders between 68 and 73C, you get a 3-year-warranty and an expectation of up to 10 years lifespan. Of course, the processor might occasionally heat up to 55C for normal usage. But then, you would expect a weighted average of temperatures to affect the lifespan taken over time, weighted for duration.

If I stress test my Skylake @ 4.7 with LinX, I'll have temperatures approaching 80C max and 76C average for the duration of the test. When I play my mildly-taxing games, the processor cores barely exceed 45C.

I'm not even arguing a point with this, but merely a way of looking at the temperature issue.

Of course, the wider swings between idle and load temperatures may be a factor, but the processors are made to run 24/7 at the TCASE spec for an upper limit. Yet, no OEM likely produces a system that comes even close to that, whatever mainstreamers or gamers choose to do with their systems -- exclusive of overclocking.

I would either call this my opinion, or an impression, but it makes sense to me.

 

[StefanR5R]

I'm planning out a dual Xeon build for the next year. You got any pics of those coolers clearing the ram? My research shows that fan barely clears a plain memory module without a heatsink. I'd prefer to go air for this build and was thinking about using a pair of NH-D15's.

I have got NH-D15S with only one fan in the middle, hence no RAM conflict.
(The symmetric NH-D15 wouldn't fit on my board anyway, at least not the way I want them: Blowing upwards in parallel, not blowing front-to-back in series.)

Spoiler: pics of the dual-Xeon + Noctua NH-D15S + Supermicro X10DAX + Corsair 400Q

As you can see, the two coolers only fit because they are asymmetric, and the left one leans to the left, and the right one to the right.

Cooling this system works very well because the top exhaust is not obstructed by a desk or shelf, hence the exhaust air is practically not circulating back into the front and back intakes.

The case provides only standoffs for ATX boards. To mount the SSI-EEB motherboard, I had to drill additional holes through the backplate and put aluminum rails at the top.

 

[bigboxes]

Spoiler: pics of the dual-Xeon + Noctua NH-D15S + Supermicro X10DAX + Corsair 400Q

As you can see, the two coolers only fit because they are asymmetric, and the left one leans to the left, and the right one to the right.

Cooling this system works very well because the top exhaust is not obstructed by a desk or shelf, hence the exhaust air is practically not circulating back into the front and back intakes.

The case provides only standoffs for ATX boards. To mount the SSI-EEB motherboard, I had to drill additional holes through the backplate and put aluminum rails at the top.

Thanks for sharing. Nice build. I'm still confused as to how I want to cool them and their configuration. Something to think about.