An Enthusiast Review of CoolClouds' Revolutionary CPU Cooler Prototype

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456

Introduction



Hello everyone, and thank you for stopping by! Recently, I had a very unique and exciting opportunity present itself. Today, I would like to share the details of the experience with my fellow enthusiasts!

Living in northern California comes with many perks (like amazing weather), but one of my favorites is that many tech companies call it home! As a result, I've had the privilege of visiting the offices of some amazing tech companies like Intel, AMD, Nvidia, ASUS, Antec, etc. In addition to the large and established tech companies, many hungry tech start ups can be found here as well.

By chance, I noticed that a tech start up by the name of CoolClouds was looking for a computer hardware enthusiast in the northern California area. They had developed an early prototype of a CPU cooler that incorporated the technology and techniques they had been using in their business cooling products. CoolClouds explained that their typical business involved designing custom cooling solutions for companies for electronics such as servers, telecommunication devices, and military equipment applications. They were not very familiar with the aftermarket CPU cooling environment and were looking for feedback.

Fast forward a short while and I found myself with an early CoolClouds CPU Cooler Prototype in my hands, hot off the production line! CoolClouds was generous enough to give me hands-on experience with their prototype and allow me to run a slew of enthusiast inspired tests on it. Below, you will find the results!






System Setup



Test Bed:

  • Motherboard: ASUS Maximus VI Impact
  • CPU: Intel Core i7 4770k
  • Memory: 8 GB (2x4GB) Samsung DDR3 1600 "Magic Ram"
  • Video: Integrated
  • Hard Drive: Samsung 840 Pro 256GB SSD
  • OS: Windows 8.1 64 Bit
  • Thermal Paste: Noctua NT-H1


Test Methods


  • CPU: CPU temperature will be monitored in windows using AIDA64 Extreme Version 4.30.2900
  • GPU: The integrated GPU in the Intel i7 4770k will be used, eliminating the heat of the discrete GPU as a variable.
  • CASE: The case used for this first round of testing is the Antec VSK-4000E Mid Tower Case
  • Thermal Paste: Noctua's NT-H1 thermal paste was used for a few key reasons: It's well known in the community for its performance and for not having an observable curing time, and I had a large syringe of it already =)
  • Temperatures:
    • Ambient: Ambient temperature will be measured throughout the test using an Exergen Infrared Thermometer. If there is any change in ambient temperature greater than .5C during the course of the test, the test will be restarted.
    • Idle: Idle temperature will be recorded after a 25 minute period of inactivity.
    • Load: Load temperature will be recorded following a 20 minute period of 100% load for air coolers and after a 30 minute period of 100% load for air/water hybrid coolers. The load will be obtained by running the AIDA64 Extreme Version 4.30.2900 System Stability Test. The values reported will represent the average temperature during the load testing period.
  • Sound: Sound levels will be measured with a Sinometer JTS-1357 Digital Sound Level Meter. A measurement will be taken at a distance of 1ft, .5m, 1m, and 4 ft. In each instance, the ambient sound level of the room will be recorded.


CoolClouds CPU Cooler Prototype


The CoolClouds CPU Cooler Prototype has a lot of familiar features to other coolers I have worked with in the past, at first glance. However, after taking it all in, it is clear that it is definitely something new, and something cool. On the one hand, I can see that it has a copper plate for contact with the CPU and I can see that the entire unit is designed to mount directly to the motherboard. But on the other hand, I see a 120mm radiator and what looks to be to be none-other than a water pump like the ones I have used in custom water cooling loops! Missing, however, are the tubes typically involved in a hybrid cooler setup. Also, the all-copper nature of the cooler is simply something to behold.

In addition to having full access to the prototype, the CoolClouds crew also made themselves available for any of my questions or other needs. They quickly brought me up to speed on the CoolClouds technology set that includes custom-designed and optimized micro-channel cold plates, powerful, miniaturized water pumps, and fine-tuned all copper radiators. With this technology set, CoolClouds has been morphing into some of the most challenging design spaces and has shown time and time again that they can provide greater cooling performance than any other current solution on the market.

In speaking with the CoolClouds team, it is clear that they are very confident in their designs and their technology. I think it was very bold of them to reach out directly to the enthusiasts and to allow one of them to pit their first prototype against the competition. Lets find out if their confidence and enthusiasm are backed up by performance, shall we? =)



Please Excuse My Rusty Photoshop Skills =)


Specifications

  • Cooler Dimensions: 165 mm x 120 mm x 16 mm
  • Weight: 900g
  • Fin Material: Copper
  • Cold Plate Material: Copper
  • Compatibility: All Modern CPU Sockets
  • Fan: Nidec-Servo Gentle Typhoon (x2)
    • Size: 120 mm x 25 mm
    • Speed: 800-1850 rpm
    • Air Flow: 58 CFM
    • Noise: 26 dB


Having designed and built computers for over 15 years, I have seen a lot of CPU coolers and their specs. What stands out to me looking at these specs is that the radiator itself is only 16mm thick. I am interested to see how this factors into performance. Additionally, I am very excited to see both copper cold plate and copper radiator! Most, if not all, of the AIO hybrid coolers out on the market use aluminum radiators. This has the effect of reducing thermal performance. Also, since the system has a mix of copper and aluminum, they are forced to use fluids other than pure, distilled water. This also has the effect of reducing thermal performance. It will be interesting to see if the all-copper construction can counter-balance the reduced thickness of the radiator.


Photos:


Here are some shots of the front, back, and base of the prototype:​













Installation and Contact:



Being a prototype, no installation manual was provided. I've installed quite a fair number of heatsinks, but honestly, the installation of the prototype was extremely straightforward and did not require much mechanical ability. First, I placed the provided backplate on the rear of the motherboard, lining it up with the socket holes. Second, I hand-tightened 4 standoffs into the backplate from the other side of the motherboard. Third, I placed the appropriate amount of thermal paste on the center of the processor heatspreader. Fourth, I placed the prototype (without fans installed) on the processor. Fifth, I installed the provided screw/spring assemblies through the four bracket holes and into the previously installed standoffs. There were absolutely no obstructions or tight spots preventing me from easily tightening the screws. As always, I took my time tightening the 4 screws moving in an "X" pattern to ensure even pressure across the CPU. I then attached the two Gentle Typhoon fans and connected the pump and the 2 fans to the appropriate motherboard headers.







After the performance testing was complete, I removed the prototype and verified that I had solid contact between the cooler and the processor. As you can see in the image, good contact was present.


With all of that out of the way, lets take a look at how it performed!


Performance Results:



Results: Temperatures


For stock testing, my Intel Core i7 4770k will be set at 3.50 GHz and 1.05 volts. For overclocked CPU testing, my CPU will be set at 4.4 GHz and 1.20 volts.

Please note that the performance chart represents degrees Celsius over the ambient temperature at the time of testing. The temperatures listed DO NOT represent direct core temperatures.



Overclock Settings




With the Intel Core i7 4770k set at 4.4 GHz and 1.20 volts, we present a worthy challenge to the established kings of the CPU cooling realm as well as to the new challenger. Claiming the top performance spot with a 1 degree Celsius lead is the CoolClouds Prototype equipped with the two fans from the Corsair H80i cooler! To me, this is fairly remarkable; the CoolClouds Prototype radiator is only 16mm thick vs the Corsair H80i radiator which is a very stout 38mm thick. In an apples to apples performance shootout with the same testbed and and same fan configuration, the CoolClouds Prototype is able to best the Corsair H80i with less than half the radiator thickness! I found this a little hard to believe, so I ran the test 3 times on both coolers to verify the results.

Coming in second and third place in thermal performance in the overclocked testing were the Corsair H80i and the Noctua NH-D14 respectively. I was fairly impressed with the Noctua NH-D14 air cooler's ability to compete with the Corsair H80i All In One water cooler. The Noctua NH-D14 is significantly larger though. One important note that I will make is that the Noctua NH-D14 is able to provide it's 3rd place cooling performance at a much quieter level than the Corsair H80i as you will see in the coming sound performance evaluation.

Being a long time enthusiast and custom loop water cooler, I wanted to see how the CoolClouds Prototype performed with the long-running champion of performance/noise radiator cooling fans (The Gentle Typhoon AP-15). If you have never had the privilege of using a Gentle Typhoon, they provide an astounding amount of cooling performance on radiators, but at a noise level that is much less than the competition. Looking at the chart, we can see that the CoolClouds Prototype equipped with a pair of GT AP-15, with its diminutive 16mm radiator, performed within 2 degrees Celsius of the Corsair H80i and the Noctua NH-D14.

The Noctua NH-D14 represents best in class performance in the air cooler category according to many enthusiasts and performance reviews alike. Likewise, the Corsair H80i represents best in class performance in the 120mm AIO cooler category. In spite of this, The CoolClouds Prototype came out swinging and demonstrated that even in such an early stage of development, it was ready to take on all challengers and bring new levels of performance and a new CPU cooler form factor to the enthusiast cooler market.


Results: Sound


Please note that the sound performance chart represents measurements from 4ft, 1m, .5m, and 1ft. All measurements were taken with the case door removed and with the decibel meter aimed directly at the CPU cooler being measured.





As I alluded to earlier in the performance testing commentary, the Corsair H80i is much louder than the Noctua NH-D14 with the fans at 12v. The measurements verify this claim as can been seen in the chart.

Subjectively, I found both the Noctua NH-D14 and the CoolClouds Prototype equipped with the Gentle Typhoons to provide a very similar acoustic profile with the sound of the air flowing through the coolers barely perceivable. On the other hand, I found the acoustic profile of the Corsair H80i to be quite loud and disruptive. It definitely made a lot of noise. In the case of the Corsair H80i, it's performance comes at the cost of a quiet computing experience.


Conclusion



The Bottom Line



First of all, I would like to thank the fine folks at CoolClouds for the opportunity to work hands-on with their first prototype and for their open-door policy regarding any questions or comments that I had throughout the experience. It is clear that they are very busy, but that they take the quality of their product and their customer's experience as the highest priority.

Early on, I had concerns that the 16mm radiator would prove too thin to provide adequate performance. As it turned out, The 16mm all-copper radiator of the CoolClouds prototype proved superior to the 38mm aluminum radiator of the Corsair H80i. As a matter of fact, the slim 16mm CoolClouds Radiator was able to provide superior performance over the gargantuan 100mm thickness of fin tower used by the Noctua NH-D14! How embarrassing!

I would also like to take a moment to talk about the form factor. In the testbed, I utilized the small but mighty ASUS ROG Maximus VI Impact. The Impact is an mITX board. In the case of the Noctua NH-D14, the cooler literally consumed the entire motherboard! I will provide images, but they simply don't do the situation justice. I appreciate the fact that the Noctua NH-D14 mounts directly to the motherboard and does not require the sacrifice of a case fan, but the cooler is just insanely massive! In the case of the Corsair H80i, the installed size was more manageable, but I was forced to sacrifice the rear exhaust of my case. This is a problem for me. In most of the systems I have built in recent years, the discrete graphics cards are by far the greatest producers of heat. For this reason, it is paramount to me that the case airflow is able to function as it was designed to aid in cooling passive motherboard components and removing the excess heat generated by the GPU(s). In the case of the CoolClouds Prototype, it mounted directly to the motherboard and installation was done! I was also able to maintain the use of my case exhaust AND I had plenty of room to work with, no special tools required.






The CoolClouds CPU Cooler Prototype, for me, was like glancing into the future of electronics cooling. It made a very big impression on me. When I look at air coolers, I see a cooling technology that has reached the limits of its performance. When I look at AIO coolers, I see case airflow design thrown out the window, and I see GPU's and CPU's fighting over case exhaust and inlet locations. When I look at the CoolClouds prototype, I see a smaller cooler with better performance! I see a cooler without compatibility issues that allows anyone to do a drop-in-replacement and explore the boundaries of their CPU's performance without thermal limitations. I see a cooling technology that can let Intel and AMD know that they can go ahead and raise the TDP (Thermal Design Power) of their desktop processors, because a new standard of cooling is in place to handle it. I look forward to seeing CoolClouds consumer products, and I think you probably would too!


The Future



During my time working with the CoolClouds team and the prototype, there was a constant line of communication. They had a lot of questions about the heat problems computer enthusiasts run into and the pro's and con's of the different solutions out there. To me, it seems that the team could never be happier than to be presented with a challenging heat problem that they can devour and design and engineer a powerful solution to. We had a lot of discussions and I gave a lot of feedback. They asked what I wanted to see, and this is what I told them:
  • Cooler
  • Quieter
  • Compatible
I told them that it was great that they could perform a few degrees cooler than the competition with less than half the radiator size, but that I'd love to see what they could do with a thicker radiator! I told them that having the best thermal performance is fantastic, but that it's even better when you can do it quietly! I told them that the last thing I want to do when I buy a cooler is to worry that it won't be compatible with my case, or that I won't be able to use memory with tall heat spreaders, or that it will block my PCIe slot!

You can find out more about CoolClouds @ http://www.coolclouds.net/


DISCUSSION:


What do you think? Got any questions for me or the CoolClouds team? Let me know! Is there anything else you would like to see for the enthusiast prototype testing? If it's something I can do, I'll give it a go as time allows!

Thanks! Hope you enjoyed!
TheHobbyist
 
Last edited:

geok1ng

2[H]4U
Joined
Oct 28, 2007
Messages
2,129
I lost interest somewhere around the "Weight: 1500g" spec.

This thing tried to combine the best of both worlds and ended up bringing up the worst of each alternative:

The whole point of watercooling is increasing clearance around the CPU slot and reducing motherboard stress tension caused by heatsink weight.

The good side of air cooling is that heat pipes are better heat carriers than copper or water and aluminum is lighter than copper. Comparing this to a TRUE copper, it is indeed 400g lighter, but adds a pump to decrease reliability.

There is no way such a beast would be cheap. One can even wonder if a waterblock made of silver+ acetal wouldn't beat its cost. ( all previous silver solutions where all silver, while cost wise using silver on the bottom is what matters).

To be truly revolutionary the inventor could try a closed loop of heatpipes with some esoteric pumping solution.
 

cybereality

Supreme [H]ardness
Joined
Mar 22, 2008
Messages
6,161
Good luck, but I was expecting to see some serious gains. As-is it looks competitive with other options but not a home run hit.
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
I lost interest somewhere around the "Weight: 1500g" spec.

This thing tried to combine the best of both worlds and ended up bringing up the worst of each alternative:

The whole point of watercooling is increasing clearance around the CPU slot and reducing motherboard stress tension caused by heatsink weight.

The good side of air cooling is that heat pipes are better heat carriers than copper or water and aluminum is lighter than copper. Comparing this to a TRUE copper, it is indeed 400g lighter, but adds a pump to decrease reliability.

There is no way such a beast would be cheap. One can even wonder if a waterblock made of silver+ acetal wouldn't beat its cost. ( all previous silver solutions where all silver, while cost wise using silver on the bottom is what matters).

To be truly revolutionary the inventor could try a closed loop of heatpipes with some esoteric pumping solution.
Hey there geok1ng!

It's too bad you checked out after seeing that weight, because there is a lot of good stuff if you continue =). Regarding the weight though, this is a hand-made early prototype, and as a result, it is much heavier than a production unit would be. The low volume of prototyping tends to result in manufacturing processes and choices that result in heavier products. Either way, I have it installed in my system right now and I have never had any concern about it's weight being enough to damage anything. With the backplate and fastening mechanism, there really isn't any danger.

I think I disagree with you about the whole point of watercooling. In my opinion, the whole point of water cooling is to use a superior cooling technique to achieve superior cooling. (and the whole point is also to have a lot of fun designing and building a custom loop =) If you take a look at the images of the three coolers mounted, the CoolClouds unit definitely has the greatest clearance around the CPU.

Regarding heat pipes being better heat carriers than copper or water: I'm going to send that question along to the engineers at CoolClouds to get their input. From my perspective though, heat pipes appear to be inferior at cooling compared to a copper microchannel coldplate and a water loop. It will be interesting to see what the experts have to say about it. =)

Regarding the cost of the cooler: The production version of this cooler is currently available on IndieGoGo for $99.00 US. I think that price is pretty much where premium coolers are at, including the MSRP for the H80i and the D14 that were in the review. A silver cooler sounds cool, but I bet graphene and other carbon tech will be the next big shift in cooling materials. (But what do I know =).

Thanks for the message and I'll follow up about the heatpipes when I hear back from the CoolClouds team!
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
Good luck, but I was expecting to see some serious gains. As-is it looks competitive with other options but not a home run hit.
Hey there cyberreality!

I do agree with you that this early prototype looks competitive, but it isn't in the "shut up and take my money" category yet. Luckily, this was just the first prototype and it served more as a proof of concept than a performance demonstration. With over 6 months of additional design and optimization, I look forward to firing up the production model on the testbed and seeing those serious gains we are both looking for!

6 months of design and optimization is a long time, and I expect to see some impressive results.

Cheers!
 

Activate: AMD

[H]ard|Gawd
Joined
Nov 6, 2004
Messages
1,988
The good side of air cooling is that heat pipes are better heat carriers than copper or water and aluminum is lighter than copper
Not true. Heatpipes are usually filled with water and work through natural convection and wicking. Using a mechanical pump means you can push a much higher volume of water at a much higher rate than a heatpipe that has to rely on converted thermal energy to power it. More volume = more heat conducting fluid being carried towards the cooler surface. Heatpipes are better carriers of heat than copper because they are water, copper is better at dissipating heat

That said, while its a cool design, but who is it for? It has more clearance around the socket than the high end massive air coolers, but not as much as a traditional CLC with tubes, and doesn't have the mounting flexibility of a tubed CLC. I can't really think of an application where this specific design is functionally better than a normal CLC in any meaningful way (since its basically a tubeless CLC) and plenty of applications where it would be worse. (e.g. an ITX case where vertical clearance is tight but there are 120mm fan holes for a normal AIO)
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
Not true. Heatpipes are usually filled with water and work through natural convection and wicking. Using a mechanical pump means you can push a much higher volume of water at a much higher rate than a heatpipe that has to rely on converted thermal energy to power it. More volume = more heat conducting fluid being carried towards the cooler surface. Heatpipes are better carriers of heat than copper because they are water, copper is better at dissipating heat

That said, while its a cool design, but who is it for? It has more clearance around the socket than the high end massive air coolers, but not as much as a traditional CLC with tubes, and doesn't have the mounting flexibility of a tubed CLC. I can't really think of an application where this specific design is functionally better than a normal CLC in any meaningful way (since its basically a tubeless CLC) and plenty of applications where it would be worse. (e.g. an ITX case where vertical clearance is tight but there are 120mm fan holes for a normal AIO)
Hey there Activate: AMD!

Thanks for your perspective on the heatpipe vs copper/water radiator situation.

Regarding who the design is for: it's definitely for standard desktop tower using enthusiasts. In my opinion, I don't see what the traditional CLC's with tubes have going for them as "mounting flexibility." I covered that a bit in my review. With this design, anyone can purchase it and drop it in their tower case without any fears of incompatibility. Traditional CLC's with tubes are not a sure fit, although cases have been trying to adapt to the changing times.

Regarding this design being functionally better than a traditional CLC with tubes: it has increased compatibility, it allows for the case inlet and exhaust fans to function as they were designed to, and it performs better. With a 16mm thick radiator, the CoolClouds design performed better than the Corsair H80i with a 32mm thick radiator. (And we are talking about the first prototype). To me that is meaningful, and why I'm excited about this tech =)

Regarding ITX with tight vertical clearance: If you have seen my ThreeZero Project thread and my Spartan thread, you know my bread and butter is the situation you describe! =) This design is clearly for tower case users, but having seen their business models in person, I know that they could blow the performance of the current "vertical" heatsinks out of the water in the future. I hope CoolClouds sees that enthusiasts are interested in seeing their technology for consumer systems so that we can get a whole range of new cooler form factors out of them =)

Cheers!
 

Activate: AMD

[H]ard|Gawd
Joined
Nov 6, 2004
Messages
1,988
Hey there Activate: AMD!

Thanks for your perspective on the heatpipe vs copper/water radiator situation.

Regarding who the design is for: it's definitely for standard desktop tower using enthusiasts. In my opinion, I don't see what the traditional CLC's with tubes have going for them as "mounting flexibility." I covered that a bit in my review. With this design, anyone can purchase it and drop it in their tower case without any fears of incompatibility. Traditional CLC's with tubes are not a sure fit, although cases have been trying to adapt to the changing times.

Regarding this design being functionally better than a traditional CLC with tubes: it has increased compatibility, it allows for the case inlet and exhaust fans to function as they were designed to, and it performs better. With a 16mm thick radiator, the CoolClouds design performed better than the Corsair H80i with a 32mm thick radiator. (And we are talking about the first prototype). To me that is meaningful, and why I'm excited about this tech =)

Regarding ITX with tight vertical clearance: If you have seen my ThreeZero Project thread and my Spartan thread, you know my bread and butter is the situation you describe! =) This design is clearly for tower case users, but having seen their business models in person, I know that they could blow the performance of the current "vertical" heatsinks out of the water in the future. I hope CoolClouds sees that enthusiasts are interested in seeing their technology for consumer systems so that we can get a whole range of new cooler form factors out of them =)

Cheers!
When I refer to mounting flexibility, I mean the ability to put the bulk of your cooling solution into multiple positions, or out of the way of the stuff on the motherboard. With tubing you can put the rad in several different spots depending on the case (back, front, top 120mm fan mounts). The niche that this solution serves is basically someone without the ability to mount a tubed system anywhere, a situation you alluded to yourself as being less likely over time. I can't argue with the cooling results, though I think it probably has more to do with the quality of the radiator assembly than anything else.
 

geok1ng

2[H]4U
Joined
Oct 28, 2007
Messages
2,129
$99 is a decent price if it can beat a Nepton 140XL. Looking at the H80i numbers, it may not beat an H75 at $75...I have no idea how they put 2 GTs on that and still sell it below $100:cool:

Heat pipes>>>>copper or water

The effective thermal conductivity varies with heat pipe length, and can approach 100,000 W/m K for long heat pipes, in comparison with approximately 400 W/m K for copper.

Weight is a great concern- we can't send a 1500g heatsink mounted, and dual/quad CPUs are impossible with such beast.

Pumps have improved, but they still fail. Heatpipes don't fail and a normal cooler of similar weight can even handle the CPU in case of a single fan fail...it is not uncommon for the kind of user that expends $99 on a CPU cooler to leave the machine running for days unattended. a dual fan tower is the most reliable solution yet.

I am glad there are news players and ideas on the cooling market.

Here my suggestion:

silver cold plate + aluminium mount and fins, add a pair of nozzles on top so we can connect it to a custom loop if we want. i would pay twice the price for a silver cold plate + hybrid cooling solution, especially if i can add 4 layers of redundancy: 2 GTs fans, 1 integrated pump, 1 external pump.:D
 

SticKx911

2[H]4U
Joined
Mar 14, 2004
Messages
2,410
I skimmed, but this is an interesting idea. The option to mount horizontally or vertically would be nice. Is it feasible to go with vapor tech instead of a pump? Since the rad is mounted to the cpu, you can take advantage of heatpipes through the rad to really make this a solid hybrid.
 

Erasmus354

[H]F Junkie
Joined
Mar 12, 2004
Messages
9,392
$99 is a decent price if it can beat a Nepton 140XL. Looking at the H80i numbers, it may not beat an H75 at $75...I have no idea how they put 2 GTs on that and still sell it below $100:cool:

Heat pipes>>>>copper or water

The effective thermal conductivity varies with heat pipe length, and can approach 100,000 W/m K for long heat pipes, in comparison with approximately 400 W/m K for copper.

Weight is a great concern- we can't send a 1500g heatsink mounted, and dual/quad CPUs are impossible with such beast.

Pumps have improved, but they still fail. Heatpipes don't fail and a normal cooler of similar weight can even handle the CPU in case of a single fan fail...it is not uncommon for the kind of user that expends $99 on a CPU cooler to leave the machine running for days unattended. a dual fan tower is the most reliable solution yet.

I am glad there are news players and ideas on the cooling market.

Here my suggestion:

silver cold plate + aluminium mount and fins, add a pair of nozzles on top so we can connect it to a custom loop if we want. i would pay twice the price for a silver cold plate + hybrid cooling solution, especially if i can add 4 layers of redundancy: 2 GTs fans, 1 integrated pump, 1 external pump.:D
Silver is only marginally better for a cold plate, it tarnishes, and is much much more expensive.
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
When I refer to mounting flexibility, I mean the ability to put the bulk of your cooling solution into multiple positions, or out of the way of the stuff on the motherboard. With tubing you can put the rad in several different spots depending on the case (back, front, top 120mm fan mounts). The niche that this solution serves is basically someone without the ability to mount a tubed system anywhere, a situation you alluded to yourself as being less likely over time. I can't argue with the cooling results, though I think it probably has more to do with the quality of the radiator assembly than anything else.
If these guys are going to make quality radiators that give superior performance, I'm interested =)

For me, the form factor is not a niche, its the standard that computers started with. I've been designing and building systems for over 15 years now, so maybe I'm just Old School =) To me, it is AIO coolers that are niche and enthusiasts have had to adapt to them in order to benefit from the better cooling they can provide. My2Cents :cool:

Cheers!
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
$99 is a decent price if it can beat a Nepton 140XL. Looking at the H80i numbers, it may not beat an H75 at $75...I have no idea how they put 2 GTs on that and still sell it below $100:cool:

Heat pipes>>>>copper or water

The effective thermal conductivity varies with heat pipe length, and can approach 100,000 W/m K for long heat pipes, in comparison with approximately 400 W/m K for copper.

Weight is a great concern- we can't send a 1500g heatsink mounted, and dual/quad CPUs are impossible with such beast.

Pumps have improved, but they still fail. Heatpipes don't fail and a normal cooler of similar weight can even handle the CPU in case of a single fan fail...it is not uncommon for the kind of user that expends $99 on a CPU cooler to leave the machine running for days unattended. a dual fan tower is the most reliable solution yet.

I am glad there are news players and ideas on the cooling market.

Here my suggestion:

silver cold plate + aluminium mount and fins, add a pair of nozzles on top so we can connect it to a custom loop if we want. i would pay twice the price for a silver cold plate + hybrid cooling solution, especially if i can add 4 layers of redundancy: 2 GTs fans, 1 integrated pump, 1 external pump.:D
Hey hey geok1ng!

I think the price point is pretty killer as well, considering what you are getting! :cool: Performance wise, the Corsair H80i is the highest performing 120mm AIO cooler by my understanding. If the 16mm thick radiator CoolClouds Prototype bested the H80i with a 38mm thick radiator, I think the production unit will be smoking it!

Hahaha! I looked it up on Wikipedia as well! Let me ask you this though geok1ing: If the heat pipe solution is so much better, then why does the CoolClouds prototype and the Corsair H80i perform better? Not only that, but the Noctua NH-D14 has 2 fin towers with 100mm of thickness vs the 16mm for the CoolClouds and the 38mm of the Corsair H80i.

Either way, the CoolClouds engineers should be getting back to me shortly with their technical answer.

I agree with you about the weight... If you send, or move a system with a Noctua NH-D14 installed, it's going to trash your machine. I don't use Dual core systems, so I hadn't thought about that at all. Really though, that weight is for a hand-made prototype that uses manufacturing techniques and materials that won't be used in the finished product. I'm going to get an updated number concerning the production units weight from the guys at CoolClouds.

Regarding pump failures... most modern CPU's throttle or shut down rather than burn up. Also, Even with the pump failing, I think as long as the fans were blowing this unit would continue to be able to adequately cool a stock or lightly overclocked processor. I'm going to run some more tests to check that idea out =)

I share your excitement for new ideas and new players in the cooling market! :cool:

Your ideas are unique for sure! Hehehe!
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
I skimmed, but this is an interesting idea. The option to mount horizontally or vertically would be nice. Is it feasible to go with vapor tech instead of a pump? Since the rad is mounted to the cpu, you can take advantage of heatpipes through the rad to really make this a solid hybrid.
Hi there SticKx911!

This cooler is designed to use copper/water/pump/radiator technology. These guys seem to know about all of the options out there, so perhaps they would have something like that in a future product :D

Cheers!
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
Silver is only marginally better for a cold plate, it tarnishes, and is much much more expensive.
Hey there Erasmus354! Thanks for stopping by and reading my amateur review =) I hope you enjoyed! You are right on all counts about silver! Plus, it is needed to kill werewolves.

Cheers!
 

geok1ng

2[H]4U
Joined
Oct 28, 2007
Messages
2,129
I think the price point is pretty killer as well, considering what you are getting! :cool: Performance wise, the Corsair H80i is the highest performing 120mm AIO cooler by my understanding. If the 16mm thick radiator CoolClouds Prototype bested the H80i with a 38mm thick radiator, I think the production unit will be smoking it!

Hahaha! I looked it up on Wikipedia as well! Let me ask you this though geok1ing: If the heat pipe solution is so much better, then why does the CoolClouds prototype and the Corsair H80i perform better? Not only that, but the Noctua NH-D14 has 2 fin towers with 100mm of thickness vs the 16mm for the CoolClouds and the 38mm of the Corsair H80i.
Actually the $75 H75 beats the H80i, and both lose to the Noctua NH-D14, Noctua's solutions also is quieter at load

Having Gentle Typhoons is a great boom for this hybrid cooler, and coupled with copper radiator explaims some of the performance gains, but unless the project somehow makes it double duty as a heat pipe, a pump failure could be catastrophic.

Silver is about 10 times the price of copper for 10% better thermals, but silver waterblocks are only ~3-6x the price of copper solutions (milling silver is easier; 0.999 silver is extremely ductile and even 0.925 sterling silver is a more friendly material to work than copper), for up to 2ºC better temps. Hence giving up the copper radiator to decrease weight can be compensated by a silver cold plate. Same performance, less weight, more epeen.

Silver tarnish is a non issue when silver is in direct contact with metals that behave as sacrificial anodes, like aluminum or steel from the mounting system. No commercial silver waterblock suffered from tarnishing.
 
Last edited:

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
Actually the $75 H75 beats the H80i, and both lose to the Noctua NH-D14, Noctua's solutions also is quieter at load

Having Gentle Typhoons is a great boom for this hybrid cooler, and coupled with copper radiator explaims some of the performance gains, but unless the project somehow makes it double duty as a heat pipe, a pump failure could be catastrophic.

Silver is about 10 times the price of copper for 10% better thermals, but silver waterblocks are only ~3-6x the price of copper solutions (milling silver is easier; 0.999 silver is extremely ductile and even 0.925 sterling silver is a more friendly material to work than copper), for up to 2ºC better temps. Hence giving up the copper radiator to decrease weight can be compensated by a silver cold plate. Same performance, less weight, more epeen.

Silver tarnish is a non issue when silver is in direct contact with metals that behave as sacrificial anodes, like aluminum or steel from the mounting system. No commercial silver waterblock suffered from tarnishing.
Hello geok1ng, champion of silver! :D

If the H75 and the H80i lose to the Noctua NH-D14, then their results differ from mine. You can clearly see that the Corsair H80i had better thermal performance than the Noctua NH-D14 on my test bed. I mirrored my testing and review off of [H]ardocp's very own testing as well. I used nearly the same hardware and nearly the same testing methods.

I'm not quite sure I follow you with this pump failure being catastrophic situation.

1) All modern CPU's will throttle or turn off rather than overheat.

2) All of the current AIO coolers use pumps as well, so they would have the same problem.

3) I ran the prototype with no pump after reading your concerns. With the fans still running, the cooler still worked extremely well and more than adequate to cool the CPU at stock settings under load. I'd like to investigate this further for my own interests, so maybe I will post an update regarding fanless/pumpless operation as I get some free time =)

Silver seems pretty cool to me! I'm not an engineer though and I don't make heat sinks for a living :cool: I will pass the suggestion along to the team at CoolClouds though!

Cheers!
 

geok1ng

2[H]4U
Joined
Oct 28, 2007
Messages
2,129
I do believe that the better way of testing is something like Anandtech does: use a dump thermal load.

it is not simple to determine the pecking order, as the results vary with the thermal load. Higher thermal loads tend to favor larger radiators, while lower loads favor better fans/blocks. The fact that the H75 comes with 2 fans that i consider to deliver the best price/noise/temps balance it's $75 price tag makes it the one to be beaten.
The H80i did not bested the X40 very often in reviews, while the cheaper H75 did it more than once. Also H80i is not compatible with GPU brackets.

About silver champion: it is all about epeen. Block design and flow matters more than material used. a good copper block would beat a bad silver one. I am not really that kind of silver champion:cool:

On the other hand, 1500g IS a problem. :mad:

I can totally see people using this if it somehow becomes lighter and can be added to a custom loop.
 
Last edited:

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
The figure I was given for the weight of the prototype was 1500g. However, it must have been an over-estimate on their part because I measure about 900g with 2 fans installed. 900g is about the same as the Noctua NH-D14 for reference.
 

DraginDime

[H]ard|Gawd
Joined
Jan 12, 2012
Messages
1,427
This is a cool concept. I really like this form factor, much better than the AIO/CLC solutions we currently have on the market. As long as the performance is right at what current CLC units are doing, then I think it's hitting the mark. Also if it still cools decently without the pump, and allows you to still use your pc in case of a pump failure, that puts it on another level.
The only fault may be the weight of it. A good backplate to distribute the weight always helps a lot in that aspect, but could still be an issue for lower budget boards that are thinly constructed.

On newegg, it shows the Noctua NH-D14 weighs in at 1240g. Maybe that's package weight? :D

I will be looking forward to the final design to see how it comes out in the end.
 

Unknown-One

[H]F Junkie
Joined
Mar 5, 2005
Messages
8,903
I ran the prototype with no pump after reading your concerns. With the fans still running, the cooler still worked extremely well and more than adequate to cool the CPU at stock settings under load. I'd like to investigate this further for my own interests, so maybe I will post an update regarding fanless/pumpless operation as I get some free time =)
Passive water cooling (no pump) has been done before, it just requires careful calibration so that the fluid in the loop moves effectively via convection.

It's also HIGHLY orientation-dependent. A thermosiphon that works upright will tend to totally fail when turned 90 or 180 degrees.
 
Last edited:

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
I do believe that the better way of testing is something like Anandtech does: use a dump thermal load.

it is not simple to determine the pecking order, as the results vary with the thermal load. Higher thermal loads tend to favor larger radiators, while lower loads favor better fans/blocks. The fact that the H75 comes with 2 fans that i consider to deliver the best price/noise/temps balance it's $75 price tag makes it the one to be beaten.
The H80i did not bested the X40 very often in reviews, while the cheaper H75 did it more than once. Also H80i is not compatible with GPU brackets.

About silver champion: it is all about epeen. Block design and flow matters more than material used. a good copper block would beat a bad silver one. I am not really that kind of silver champion:cool:

On the other hand, 1500g IS a problem. :mad:

I can totally see people using this if it somehow becomes lighter and can be added to a custom loop.
I like the way Anandtech tests as well, especially with their latest AIO roundup. HOWEVER, I only prefer it as a synthetic test. Anandtech method is like a synthetic benchmark and hardocp testing is closer to what you would actually experience on an enthusiast platform (In My Humble Opinion =)

I hope they are get the weight under control as well. Mostly, I hope they bring awesome new toys into the enthusiast marketplace for me to play with =)

Thanks!
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
This is a cool concept. I really like this form factor, much better than the AIO/CLC solutions we currently have on the market. As long as the performance is right at what current CLC units are doing, then I think it's hitting the mark. Also if it still cools decently without the pump, and allows you to still use your pc in case of a pump failure, that puts it on another level.
The only fault may be the weight of it. A good backplate to distribute the weight always helps a lot in that aspect, but could still be an issue for lower budget boards that are thinly constructed.

On newegg, it shows the Noctua NH-D14 weighs in at 1240g. Maybe that's package weight? :D

I will be looking forward to the final design to see how it comes out in the end.
I'm with you on this one DraginDime! sounds like we are of same mind here. With the current AIO coolers, the tubes with water are the only path for heat to the radiator. In the case of this form factor, the heat can still travel through the metal of the unit just like a solid copper heatsink. Fan's still blowing? Then it's just a big copper heatsink =) I will run the tests when time allows to check it out... for science... =)

These guys are pretty smart, I'm sure they will address the weight without sacrificing performance. We'll see!

Passive water cooling (no pump) has been done before, it just requires careful calibration so that the fluid in the loop moves effectively via convection.

It's also HIGHLY orientation-dependent. A thermosiphon that works upright will tend to totally fail when turned 90 or 180 degrees.
Hello Unknown-One! You sound like you know what you are talking about =) This cooler would also work as a copper heatsink as well as whatever amount the passive water cooling added.

Interesting stuff! I'll let you know how the no-pump test goes!

Cheers!
 

geok1ng

2[H]4U
Joined
Oct 28, 2007
Messages
2,129
trying to sum up the market trends on cooling:

Copper radiator: old idea. performance gains where inline with the price premium, but the weigth meant that was a solution viable only to people with absolute hate towards custom water cooling.

Integrating a cooler to the CPU block: twice old idea: apogee drive could not beat the normal pumps in performance nor reliability. And AiO coolers had performance below air cooling solutions for twice the price, with HUGE reliability problems.

So the main ideas behind the prototype were used before, with none to limited market success. We want new players on the market, so we want this product to suceed.

What has been tried before that worked?

-Selling a working-out-of-the-box water loop with the option for later adding it to a custom build loop. That was last year Swiftech's H220 and 2014 H220X.

Can this be done in this product? yes.
Will it decrease performance? no.
How much will it increase B.O.M? ~$5
Has anything close to such product ever been tried before? NOPE.

When must it be ready for sales?
Around Haswell-E launch cominh October.
 

c3k

2[H]4U
Joined
Sep 8, 2007
Messages
2,169
Graph shows the cool cloud as the hottest solution, not the coolest.

Higher temp rise over ambient than the others.
 

flod

Gawd
Joined
Jul 10, 2014
Messages
948
wait so
Will they sell these with gentle typhoons? Will they sell gt's independently?
 

geok1ng

2[H]4U
Joined
Oct 28, 2007
Messages
2,129
wait so
Will they sell these with gentle typhoons? Will they sell gt's independently?
that was precisely my point about price: GTs are not only expensive but extremely hard to find, so some people would buy the product just for a reliable source of GTs.:p
 

buzzbomb

Gawd
Joined
Sep 29, 2009
Messages
594
Also, since the system has a mix of copper and aluminum, they are forced to use fluids other than pure, distilled water.
I'm sure you can't tell us exactly what the heat transfer fluid is, but this is interesting. Judging by their industrial background, can it be assumed that it's some sort of non-conductive oil? Guess we'll have to wait until someone cracks one open. :)
 

primetime

Supreme [H]ardness
Joined
Aug 17, 2005
Messages
6,621
I'm kinda curious what the temps would be if the pump was not powered on (it looks like it would still cool if the pump failed)
 

ssnyder28

2[H]4U
Joined
May 9, 2012
Messages
3,649
This looks like a cool concept and thanks for the review. The fact that an unrefined prototype competes at the top level enthuisiast cpu cooler tier is impressive. Obviously in the end its going to come down to how cheaply they can manufacture these, to me I think it'll be hard to keep them under $100 a piece and I don't think that'll be low enough to make a major splash in the high end market.
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
trying to sum up the market trends on cooling:

Copper radiator: old idea. performance gains where inline with the price premium, but the weigth meant that was a solution viable only to people with absolute hate towards custom water cooling.

Integrating a cooler to the CPU block: twice old idea: apogee drive could not beat the normal pumps in performance nor reliability. And AiO coolers had performance below air cooling solutions for twice the price, with HUGE reliability problems.

So the main ideas behind the prototype were used before, with none to limited market success. We want new players on the market, so we want this product to suceed.

What has been tried before that worked?

-Selling a working-out-of-the-box water loop with the option for later adding it to a custom build loop. That was last year Swiftech's H220 and 2014 H220X.

Can this be done in this product? yes.
Will it decrease performance? no.
How much will it increase B.O.M? ~$5
Has anything close to such product ever been tried before? NOPE.

When must it be ready for sales?
Around Haswell-E launch cominh October.
I will pass your ideas on to the team. Good thoughts, thanks for sharing!
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
Graph shows the cool cloud as the hottest solution, not the coolest.

Higher temp rise over ambient than the others.
Hello c3k! The prototype is on the chart 2x, once in the top spot and once in the last spot. In the last spot, I decided to see what the prototype could do with some Gentle Typhoon AP 15's I had laying around.

wait so
Will they sell these with gentle typhoons? Will they sell gt's independently?
Well, the prototype will never be for sale, but they are going to be selling their enthusiast cooler with GT's =) At this time, I don't think they have plans to sell the GT's independently.

that was precisely my point about price: GTs are not only expensive but extremely hard to find, so some people would buy the product just for a reliable source of GTs.:p
I would do this =) Love GT's!

Needs more TEC.
IKR! Hahaha! I always found TEC coolers awesome, but I was always afraid of the condensation.

I'm sure you can't tell us exactly what the heat transfer fluid is, but this is interesting. Judging by their industrial background, can it be assumed that it's some sort of non-conductive oil? Guess we'll have to wait until someone cracks one open. :)
It all depends on the design requirements! Secrets are fun, aren't they? =)

I'm kinda curious what the temps would be if the pump was not powered on (it looks like it would still cool if the pump failed)
Quite a few have expressed interest in this. I will be testing it as I get a chance and I will report back here!

This looks like a cool concept and thanks for the review. The fact that an unrefined prototype competes at the top level enthuisiast cpu cooler tier is impressive. Obviously in the end its going to come down to how cheaply they can manufacture these, to me I think it'll be hard to keep them under $100 a piece and I don't think that'll be low enough to make a major splash in the high end market.
For me, cooler's have 3 factors that need to balance well: price, performance, and acoustics. If any of those 3 are off, the decision to buy is harder. Honestly though, if the performance and acoustics are awesome, I am more forgiving with price =)

Where was the "ambient" temperature reading taken?
The ambient temperature reading was taken in the room the test was performed in.
 

munkle

[H]F Junkie
Joined
Jan 16, 2005
Messages
11,505
I don't get where you are getting 16mm rad from, if you look at your pics the rad is as thick as the 25mm fans.
 

doyll

[H]ard|Gawd
Joined
Feb 4, 2012
Messages
1,222
The ambient temperature reading was taken in the room the test was performed in.
Then the results have a serious flaw.

The critical air temperature is the air going into cooler, not the room air.

The room ambient is not the same as cooler intake air temp .. which usually changes with each cooler tested, what fans are used, and what the fan speed is.

Your results are not how well the coolers perform, but how well your system performs with the different coolers. ;)

Put a temperature probe 6-9cm in front of cooler intake and you will see what I mean. :D
 

TheHobbyist

Hugs Hard Johnnies [H]ard
Joined
Apr 8, 2008
Messages
456
Then the results have a serious flaw.

The critical air temperature is the air going into cooler, not the room air.

The room ambient is not the same as cooler intake air temp .. which usually changes with each cooler tested, what fans are used, and what the fan speed is.

Your results are not how well the coolers perform, but how well your system performs with the different coolers. ;)

Put a temperature probe 6-9cm in front of cooler intake and you will see what I mean. :D
Thanks for chiming in Doyll.

That's a valid concern, but honestly, I just copied the [H] way of testing.

HOWEVER, I am working on "Test Bed #2" and it would be great to take community feedback about how you would like to see coolers tested and what tests you would like to see.

So far the questions I will try to answer with "Test Bed #2" and subsequent testing are as follows:

  • How great of an effect does side pannel on vs side panel off effect the thermal performance of each cooler?
  • How do the CoolClouds prototype and the Corsair H80i perform with no fans (cooler fans or case fans) but with the pump running
  • How do the CoolClouds prototype and the Corsair H80i perform with fans, but no pump
  • How do the different cooler form factors (I.E. AIO and motherboard mounted) affect the temperature of other components in the system? (i.e. CPU temp, case temp, vrm temp, cooler intake temp, etc)
  • How do the 3 coolers perform using cooler intake temperature instead of room ambient
  • How are temperatures affected by different AIO radiator placements (i.e. placed in rear exhaust as an exhaust, placed in rear exhaust as intake, placed in front intake as intake, etc)

What do you think?
 

W.Feather

Hard DC'er for Aug 2012 / Feb 2019 / Mar 2020
Joined
Nov 10, 2009
Messages
11,431
What is the point of testing in non-realistic scenarios. I wont be running a case w/o a side, i wont run without a fan, i wont run either without a pump. You can do these test but the results are pointless, running in a hot box, yes, but the ones I just listed are literally a waste of time IMO
 

doyll

[H]ard|Gawd
Joined
Feb 4, 2012
Messages
1,222
You seem to be deviating from what this thread is about.
  • Side panel on or of has nothing to do with the coolers' ability to cool. Only the case's ability to cool. I thought this was a cooler test, not a case test.
  • Neither the CooClouds prototype or H80i are designed to be passive coolers, so why bother? It accomplishes nothing.
  • Same for running with no pump. Both coolers are designed to have pumps running.
  • Cooling effect on system components is about the case performance, not cooler performance.
  • The cooler performance is about how well cooler can cool CPU based on air temperature going into cooler / radiator. That temperature is a result of how well the case flows air and cools.
  • Again, radiator placement reflects case cooling as much or more than cooler performance.
 
Top