Hi Everyone,

Phase two begins today, getting ready to order the following items for P3D:

1. 3 - Intel Core i7-4790K Quad-Core 4.0GHz CPUs
2. 3 - ASUS MAXIMUS VII FORMULA Motherboards
3. 3 - ENERMAX Platimax EPM1000W Power Supplies
4. 3 - GeForce GTX TITAN X Video Boards
5. 3 - Corsair Vengeance Pro 16GB / DDR3 / 2400 RAM Sets
6. 3 - Antec Twelve Hundred ATX Computer Cases
7. 3 - Crucial BX100 2.5" 1TB SATA 6Gbps SSDs
8. 3 - Sony SATA 24X Burner / DVD+R DL CD/DVD Drives
9. 3 - Prolimatech CPU Cooler with 120mm Fans

Then it's on to Phase Three, three projectors and a 220* wrap around screen. I hope to have phase three finished before the end of the year, as always this will depend on saving additional funds.

I'm getting there, slowly... but surely. 

Would 3 GTX 980 Ti's be better than three TX's?

 

I thought the TX was a waste of money

Prolimatech CPU Cooler with 120mm Fans

 

Which version? There are better.

 

The Noctua NH-D15S is now available. Legendary NH-D14/15 cooling but with much improved compatibility with ram etc.

 

There is no better air cooler.

Would 3 GTX 980 Ti's be better than three TX's?

 

I thought the TX was a waste of money

 

Depends on VRAM usage.  In terms of rendering power, they're nearly identical.  Double the VRAM allows you to run higher resolutions/AA though.  If he's running 4k or multi-monitor having 12GB instead of 6GB could be useful.  

 

Prolimatech CPU Cooler with 120mm Fans

 

Which version? There are better.

 

The Noctua NH-D15S is now available. Legendary NH-D14/15 cooling but with much improved compatibility with ram etc.

 

There is no better air cooler.

 

I wouldn't spend this much money on hardware and use air cooling.  Even closed loop, off-the-shelf systems are better and quieter than any air cooler you can buy.  Besides, if you buy a tower system and stand it up, then hang a huge hunk of metal off the CPU socket at a 90 degree angle you're putting lots of stress on the motherboard and *will* warp the board over time.  Water coolers solve this issue.  

Hi,

 

Each video board will run its own projector at 1920x1200. As for the large heat sink warping the M/B, not a chance with small silver chain link necklaces holding most of the weight clipped to the roof of the computer <- my own invention :smile: .

 

Since I am not overclocking (running @ 4GHz / no need to overclock), frames are over 100 now on my "ONE" Titan Black attached to two 42" HDTVs at a resolution of 3840x1080. Why would I want the additional cost AND maintenance involved in water cooling?... Been there / Done that... running P3D on one top of the line, state of the art computer, overclocked to the hilt chasing hardware release after hardware release, those days are over... I'll be running P3D on six I7 4790K computers with three Titan X video boards. As for the additional sound generated by the six computers, that's fine... it will sound more like an avionics bay.

I wouldn't spend this much money on hardware and use air cooling. Even closed loop, off-the-shelf systems are better and quieter than any air cooler you can buy. Besides, if you buy a tower system and stand it up, then hang a huge hunk of metal off the CPU socket at a 90 degree angle you're putting lots of stress on the motherboard and *will* warp the board over time. Water coolers solve this issue.

 

 

Sorry, but that's wrong, on both counts.

 

And I base this on personal experience.

 

My NH-D14 has been on three builds now, and I can tell you that the cooler is whisper quiet. If you look at cooler reviews that include acoustics, AIO water coolers and the D14/15, you will see that it beats the AIO water coolers for acoustics routinely. The D14 is famous for being quiet. And as far as AIO coolers, many of them are infamous for noise. The Kraken for example!

 

As for cooling it does a fantastic job. Each one of my builds that has included the D14 have been overclocked with ease. Yes, full blown loops will be cooler and ye, some AIO water coolers, like the excellent H110, will outperform it for cooling, But in many cases we don't need the extra few degrees.

 

Even closed loop, off-the-shelf systems are better and quieter than any air cooler you can buy.

 

That is utterly wrong. :smile: many of the AIO coolers don't at all beat the D14/D15, only the best do. And in many cases the AIO's cheat by using much higher rpm fans than the D14, and thus nosier.

 

 

H110i 44 degrees. NH-D14 47 degrees. Barely any difference. And you will note that the H100i is at full fan rpm, [2456 rpm] whereas the D14 [1276 rpm] beats it acoustically with ease.

 

 

http://www.tomshardware.com/reviews/nepton-280l-tundra-td02-water3.0-pro-reserator3-max,3607-11.html

 

Here, H110 36 degrees above ambient. NH-D14 38 degrees above ambient! 1155 socket. A mere two degrees better.

 

http://www.bit-tech.net/hardware/cooling/2014/05/01/noctua-nh-d15-review/2

 

Results will of course vary in the reviews, but I'm sure you get the picture. The point is that many AIO's are indeed inferior, and even the best AIO's aren't miles ahead. And noise and the D14... nope, super quiet!

 

Besides, if you buy a tower system and stand it up, then hang a huge hunk of metal off the CPU socket at a 90 degree angle you're putting lots of stress on the motherboard and *will* warp the board over time.

 

 

Nope, it wont. as I say, my D14 has been on three builds. No warping whatsoever.

 

When I first purchased the D14, I too was a little concerned about the weight. So I designed and built a support bracket. However... as soon as the cooler was fitted, it immediatly became apparent to me that my support bracket was totally unrequired, so I threw it in the bin.

 

Noctua's SecuFirm mounting system is superb, fully supports the cooler with ease. This is what Noctua have to say, and my first hand experience leads me to agree with them...

 

 

Is the high weight dangerous for the CPU or socket?

No. Noctua coolers possess an extremely reliable SecuFirm™ mounting system. Thanks to the screw connection with the backplate on the rear side of the motherboard, the exceedance of the weight recommendations by Intel and AMD common among high-end coolers is completely unobjectionable.

 

 

Furthermore, you won't find any confirmed issues with the D14/15 damaging motherboards anywhere on the internet. Noctua's engineers designed the system to avoid such issues. There are a vast number of PC's that have utilised the D14/D15 over the past few years, motherboards aren't dropping like flies.

 

I hope I haven't come across as sounding like a Noctua fanatic, I'm not, but this "it will bend and damage the motherboard" stuff is a fallacy, and given that the D14 is famous for low noise I have to speak up, just in case such comments put a potential buyer off needlessly.

From Bit-tech, re the NH-D15...

 

 

We had our concerns that the NH-D15 simply wouldn't live up to its hefty price tag - after all, many of the top-end all-in-one liquid coolers cost about the same. Thankfully, it managed some pretty shocking numbers. In dual fan mode it matched Corsair's Hydro H105 and the Phanteks PH-TC14PE in our LGA1155 test system with a delta T of 38°C, and this only rose 1°C when we removed the front fan too. In fact, even using both fans connected using the low noise adaptors resulted in a delta T of just 40°C, which is the same as Corsair's Hydro H75 and slightly better than Antec's Kühler H20 950 on its extreme profile. Needless to say the latter was a huge amount noisier too.

 

 

The NH-D15 might seem expensive, but for LGA2011 owners especially, it performs as well as all the top-end all-in-one liquid coolers out there so in cooling terms it justifies its price. Even on LGA1155, which is easier to tame, it's still extremely potent, matching the 4th best result on test. Perhaps the biggest plus point for those looking to get rid of a small, whiny cooler in time for warmer summer months, the NH-D15 is also very quiet, especially so when used with the included low noise adaptors.

http://www.bit-tech.net/hardware/cooling/2014/05/01/noctua-nh-d15-review/3

 

 

The mini review below says it all. The user kept all variables the same, same fans on the H110 and same fans on the D15, same ambient. When you do so, you get a more accurate picture of the true difference between AIO's and air cooling.

 

In this case, the D15 was 6-8 degrees cooler than the Corsair H110!

 

Don't be fooled by the fan variable chaps!

 

 

http://www.overclock.net/t/1519715/mini-review-noctua-nh-d15-vs-corsair-h110

 

 

Here the D15 doing amazingly well against the Kraken. and far more quietly.

 

http://www.tomshardware.com/reviews/zalman-reserator-3-max-dual-nzxt-kraken-x61-noctua-nh-d15,4000-5.html

 

 

Apologies for hijacking the thread, but hopefully the information will be useful to the OP.

Let me preface my comments by saying this: the following post is intended to be purely informational.  I am not attacking you in any way, so please do not take my rebuttal of your analysis personally.  

 

The first set of tests @ Tom's and Bit-tech show better results with the water coolers.  You can downplay the difference but there is a difference.  The second set of tests @ Tom's (your second post) are using low fan speeds to produce equivalent noise levels, and that's certainly one way to test but it's not accounting for the design principles of the products.  It would be like taking 2 cars, one that requires premium gas the other that runs on regular unleaded and testing them both on regular unleaded.  In this scenario you're artificially limiting the performance of the car that needs premium (such cars can run on lower octane gas, the ECU just de-tunes the motor by pulling timing).  In such a scenario we would say this is obviously a flawed testing method, in that the data it produces would only be useful to answer the question "how does the car which requires premium gas perform when it is given the wrong gasoline?"  

 

So, onto the design of these coolers.  The large tower style HSF coolers have relatively many fins with large surface area attached to multiple heat pipes to dissipate heat.  These designs have relatively low mass obstructing air flow, whereas radiators featured in water cooling adhere to different design principles which end up making them much more dense, with far more mass in the way of the fans.  If you compare footprint of the HSF fin array to that of a radiator, it should be obvious that the radiator needs to pack more, smaller fins into the same or even smaller footprint in order to achieve the same or greater surface area.  The Tom's test shows the higher speed fans on the water coolers outperforming the air cooler, even when it is paired with high speed fans.  Also, 8 degrees ain't nothing to sneeze at.  Overclock the chips further than the meager OCs featured in those tests and the water cooler will be able to hold the heat at bay for longer.  You may get an extra 100MHz of OC headroom as a result.  

 

I've used many, many cooling solutions over the years, in my own systems and in customer's systems as a PC technician/"OEM" system builder (I have designed, assembled, and sold many custom systems) for the past decade and a half.  Cooling is a particular field of interest of mine and I used to constantly "chase the dragon" in search of the best air or AIO cooling until I decided to just go full-on custom water cooling several years ago.  The extra cooling capacity afforded by "real" water cooling adds somewhere in the range of 200MHz to your top-end OC, in my experience.  This is based on extensive testing during the overclocking and stability testing phase of my Ivy Bridge system (3770k) purchased in 2012.  When I first built the system, I did so with a top-end dual-fan AIO water cooler and was able to achieve maximum OC of 4.7GHz, with temperatures that would occasionally cause thermal throttling.  That is *after* de-lidding, btw.  Prior to that the chip could only do 4.5-4.6GHz and would instantly achieve maximum temperature under stress-testing.  I decided to switch to "real" water cooling and have never looked back since.  That chip was able to run at 4.9GHz as a result, and with a lot more voltage too so the heat being generated was far more than the previous 4.7GHz configuration.

 

I also water cool my video card(s) and see outrageous temperature reductions there.  GPUs are physically much larger than CPUs and their dies are exposed directly to the cooler's mating surface so thermal transfer/conductivity is much greater in this application.  I currently have a GTX 980 Ti running @ 1547/8000 clocks 24x7 and the maximum temperature I see on that chip while stress-testing is 38 degrees C.  That's in an non-air-conditioned room in the Summer, with ambient temps in the upper 20's C.  Last Winter I decided to upgrade to X99/5960x with 3x GTX 980s (this was before the 980 Ti was available) and water cool all of it.  Since it was Winter the ambient temps in my office actually got down to the upper teens so the maximum temperature I ever saw with 3x 980s running stress-testing was 32 degrees C.  Air cooled Geforces run in the 80's, BTW.  That's up to a 50 degree temperature drop.  Now, you can say temperature doesn't matter, but when you're talking about differences that large you better believe it makes a difference in both OC potential and power draw.  If you don't care about those things that's fine, but they matter when you're looking for the best performance and you have to pay the power bill for a system like that

 

Anyway, this is all academic since the OP has stated his intentions (and re-stated them) and will continue as planned.  

Hi,

 

I like what I see in the Noctua NH-D15, did you have any issues fitting the cooler and the RAM modules getting in the way?.

Quote

The first set of tests @ Tom's and Bit-tech show better results with the water coolers. You can downplay the difference but there is a difference.

 

Yep there is a difference, didn't say there wasn't, in fact I said...

 

Quote

Yes, full blown loops will be cooler and yes, some AIO water coolers, like the excellent H110, will outperform it for cooling,

 

Don't forget, you said...

 

Quote

Even closed loop, off-the-shelf systems are better and quieter than any air cooler you can buy.

You refered to "closed loop" systems in general. Which includes a multitude of AIO coolers that are most certainty inferior to the D14/15. AIO coolers "in general" aren't better and quieter than any air cooler at all.

 

 

Quote

we would say this is obviously a flawed testing method

 

Then look at the review of the H110 compared to the D15 where the fan variable was eliminated! Same ambient temp, same fans. The D15 beat the H110 by 6-8 degrees. Manufacturers of AIO water coolers just love whacking super high RPM noisy fans on their radiators in order to imply their cooler is now the best.

 

 

Quote

The Tom's test shows the higher speed fans on the water coolers outperforming the air cooler, even when it is paired with high speed fans. Also, 8 degrees ain't nothing to sneeze at. Overclock the chips further than the meager OCs featured in those tests and the water cooler will be able to hold the heat at bay for longer. You may get an extra 100MHz of OC headroom as a result

 

I don't disagree. But it's the best AIO water coolers that beat the D14/D15, not AIO coolers in general, and when you eliminate the fan variable the difference is smaller than many would realise. And in some tests, for example the one I posted, the D14/15 actually triumphs.

 

Your initial comments made it sound as if any AIO cooler would beat any air cooler which is clearly wrong. You also wrongly labelled the D14/15 as nosier than AIO coolers when in reality the opposite is true. You also propagated the fallacy that the D14/15 warps motherboards and you made that statement definitively. The only issue I've seen regarding this related to idiots attempting to ship their PC's half way across America with a D14 still attached. Clearly a silly thing to do.

 

What most of us desire is a balance between cooling and acoustics and the Noctua cooler shouldn't be written off as an inferior product to AIO coolers. I was reading a forum this morning where an owner of an H105 ditched the cooler in favour of the D15 because he was sick of the fan and pump noise.

 

Then we have the danger of leaks to contend with. Unfortunately manufactures don't release the data concerning just how many AIO coolers leak so we're left guessing in this regard. I suspect it's not many. However, air coolers can't leak, have no moving parts except the fan and offer many years of unimpeded cooling. And when it comes to the D15, it cools an overclocked CPU very adequately indeed and very quietly. So lets not right it off as a cooler that in no way should be fitted to a high end system, as you did previously. I understand that you are someone that likes to push performance to the max, but most of us are somewhat short of that extreme.

Hi,

 

I like what I see in the Noctua NH-D15, did you have any issues fitting the cooler and the RAM modules getting in the way?.

 

 

For me no issues at all with the D14. I have Corsair Dominator sticks, the old dominators with removable top fins. Just removed them and the RAM fitted nicely. The newer D15 has been redesigned so that RAM with even taller heat spreaders can be fitted, but of course this is one disadvantage of large air coolers that needs consideration. Of course, super tall heat spreaders on RAM are just a gimmick.

 

For my next build, I'll probably be going for the D15S.

 

The D15S now has an offset design to accommodate  smaller form factor motherboards. The top PCIe slot is now clear on those motherboards. In addition, there's now just one fan.

 

I know what you're thinking, how does that affect cooling. Well actually barely at all. Makes just 2 degrees difference. Seems the second fan on the D15 isn't really required.

 

The D15S has a Noctua 150mm fan in the centre. That's one of the disadvantages of AIO coolers of course, not much airflow from the cooler over the VRM's. The 150 fan on the D15/15S generates a healthy gust across the motherboard as a result of the fan being larger than the cooler by design.

 

RAM clearance is now 65mm.

 

Excellent RAM compatibility

Thanks to its recessed lower fins, the NH-D15S provides 65mm clearance for tall memory heatsinks, making it compatible with most high end modules on the market. As the lower fins are recessed on both sides of the heatsink, compatibility is excellent not only on LGA115x and AMD, but also on LGA2011 where there are RAM slots on both sides of the socket.

 

 

 

Another example of why we shouldn't right off the D15 for high end builds. Not a huge overclock, but beating the H110 and the rest of the AIO coolers. All with the same fans.

 

 

Edit: this post has been heavily edited to comply with quotation limits imposed by the forum's software.  I originally had 8 quote blocks but the maximum, apparently, is 5.  Also, when attempting to split into 2 separate posts the forum software automatically attempts to merge the posts (and fails) due to its own limitation on quotes.  If the following isn't perfectly clear, please ask for clarification.

 

 You refered to "closed loop" systems in general. Which includes a multitude of AIO coolers that are most certainty inferior to the D14/15. AIO coolers "in general" aren't better and quieter than any air cooler at all.

 

This is a quite semantic argument you're making.  If you're saying that out of hundreds and hundreds of air coolers on the market, one or two models can beat some AIO models (but not the top ones unless you introduce needless limitations on the AIO) that doesn't exactly invalidate my general principle, it's a simple rare exception.  The rule still holds.  Water has much higher thermal conductivity than air.  That's physics 101.   

 

 

 

I don't disagree. But it's the best AIO water coolers that beat the D14/D15, not AIO coolers in general, and when you eliminate the fan variable the difference is smaller than many would realise. And in some tests, for example the one I posted, the D14/15 actually triumphs.

 

You're not being specific enough here.  Which tests?  You posted several links, be specific.  In most of the tests the top-tier AIO coolers have the best performance, so if you've got 1 out of 5 tests that shows a "win" for your argument, I'd hardly call that a win...  What do you mean by "when you eliminate the fan variable?"  Do you mean run the product out of spec and drastically lower its performance?  That's not a matter of leveling the playing field, it's an intentional limitation of the product's capabilities.  And I've explained precisely why this is the case in my last post but you didn't quote that section, so that can only mean that you either: 

1) didn't read it

2) agree with it and if that's the case, we shouldn't be having this particular discussion

 

I didn't specifically say "the D14/15 is louder than AIO coolers" I made an (obviously) general statement.  Again, just because you can find a singular exception to a rule doesn't disprove the rule.
 

 

Your initial comments made it sound as if any AIO cooler would beat any air cooler which is clearly wrong. You also wrongly labelled the D14/15 as nosier than AIO coolers when in reality the opposite is true.  You also propagated the fallacy that the D14/15 warps motherboards and you made that statement definitively. The only issue I've seen regarding this related to idiots attempting to ship their PC's half way across America with a D14 still attached. Clearly a silly thing to do.

 

 

Again, I made no specific claims about the D14/15 warping motherboards, I attributed this problem to large HSFs in general.  I have personal experience with this, I have seen it happen on a number of motherboards over the years in my work as a PC technician.  An average user is not likely to even notice such a thing.  Have you ever examined a motherboard for evidence of warping, specifically?   

 

 

What most of us desire is a balance between cooling and acoustics and the Noctua cooler shouldn't be written off as an inferior product to AIO coolers. I was reading a forum this morning where an owner of an H105 ditched the cooler in favour of the D15 because he was sick of the fan and pump noise.

 

 

"most of us"?  Who are you speaking for here?  You make it sound like AIO coolers are DC8s and your HSF is a Tesla.  Noise quality is totally subjective.  Let's be honest about that noise quality, we're talking about (at most) 2000 RPM fans here.  We're not living in the days of 7000 RPM Delta fans any more.  Where a 1000 RPM fan might be barely or not at all noticeable to most users, a 2000 RPM fan of the same size and general design is not suddenly an unbearable noise, it just crosses the threshold into noticeability for a lot of people though.  I'm not going to speculate on return rates of AIO coolers due to failure, as you say, that information isn't made public.  AIO systems can be returned due to a fault in manufacturing (usually an air bubble) or a subjective evaluation by the user. 
 

Then we have the danger of leaks to contend with. Unfortunately manufactures don't release the data concerning just how many AIO coolers leak so we're left guessing in this regard. I suspect it's not many.  

However, air coolers can't leak, have no moving parts except the fan and offer many years of unimpeded cooling. And when it comes to the D15, it cools an overclocked CPU very adequately indeed and very quietly. So lets not right it off as a cooler that in no way should be fitted to a high end system, as you did previously. I understand that you are someone that likes to push performance to the max, but most of us are somewhat short of that extreme.

 

 

I've seen 3 (out of hundreds) of AIOs that have leaked since they first started popping up in custom machines in the late 2000's.  You're reading far too much into my comments.  What I said was "I wouldn't spend this much money on hardware and use air cooling."  That indicates a personal preference.  I haven't said a single negative remark about your choice in coolers, only stated my own opinion and listed reasons why I believe water is superior.  

Hi Guys,

 

My three Super Clocked Titan X's were delivered today, on a Sunday... Sweet. Think I'm going to run these in my setup now and replace my Titan Black, when the new computers are built and the wrap is done, then I'll move them to their forever home.

 

Nice cards.  I hope they last you a long time.  Especially when you spend that kind of money.