Zalman CNPS10X Extreme Cooler Review

“Strong, handsome, face of a leader – does Arthur look like a king or what?” – ©”Shrek the Third” by DreamWorks.

Today we are going to talk about a new cooling solution from the Korean Zalman Company. I have to admit that I am very happy that Zalman has finally moved away from the radial heatsink design that was first introduced back in the far away 2005 in their Zalman CNPS9500 LED solution. At that time this cooler really offered unprecedented efficiency, but its copper glitter quickly lost its luster against the background of overwhelmingly effective Scythe Ninja, Thermaltake Big Typhoon and the whole lineup of other super-coolers following shortly after. The attempt to regain the market leadership in 2006 with the launch of Zalman CNPS9700 LED and its modifications turned out only partially successful. And the new Zalman CNPS9900 LED wasn’t quite at home among other numerous super-coolers. So, it looked like not only Zalman fans but also the company engineers and marketing specialists grew pretty tired of this state of things, because they introduced a completely new solution using classical tower-design. It was Zalman CNPS10X Extreme. It doesn’t have anything totally innovative about it, but we still have to dwell on a few very interesting solutions implemented in this cooler. But let’s start from the very beginning.

Package and Accessories

The new cooler is shipped in a box of Zalman’s traditional style and design. There is a round cut-out window in the front of it revealing the cooler fan and part of its heatsink.

All key features of Zalman CNPS10X Extreme are listed on the back of the box and its technical specifications are on one of the sides of it. Inside a thick cardboard outer box there is a clear molded plastic casing holding the cooler very securely inside. The accessories are placed right above it:

  • A plastic frame and a steel plate for LGA1366 mainboards;
  • A plastic frame and a steel plate for LGA1156/775 mainboards;
  • Backplate for LGA775 mainboards;
  • A set of retention screws and washers for LGA775 mainboards;
  • A steel plate for Socket 754/939/940/AM2(+)/AM3 mainboards;
  • A set of retention screws and plastic bushes;
  • Power cable for the rotation speed controller;
  • Zalman logo sticker;
  • Installation instructions in two languages;
  • Zalman ZM-STG2 thermal compound.

Of course, this is new Zalman thermal compound that came to replace not very efficient ZM-STG1. ZM-STG2 thermal compound is of gray color, thick and viscous consistency. The claimed heat conductivity is 4.2 W/(m·K), which is 3.5 (!) times higher than the same spec of Zalman ZM-STG1: 1.2 W/(m·K). We are not going to test the new thermal paste within our today’s test session. It will be included into our upcoming roundup of thermal compounds that we are currently collecting the samples for.

Here I would also like to add that Zalman CNPS10X Extreme is made in Korea and is priced around $60.

Design and Functionality

As we have already mentioned in the introduction to our today’s article, Zalman CNPS10X Extreme is built following classical tower design. When you get the first feel of it, you can’t help noticing that it is pretty heavy: 920 g is considered quite a lot even according to today’s standards. Let’s take a closer look at this new solution:

1 zalman cnps10x design
2 zalman cnps10x design
3 zalman cnps10x design
4 zalman cnps10x design

Fully nickel-plated heatsink consists of 56 aluminum plates, each 0.45 mm thick that are spaced at 1.7 mm apart from one another. The heatsink array sits on five copper heatpipes with 6 mm diameter. According to the official specifications, the effective heatsink surface measures 8,544 cm2. The heatsink plates array is topped with a plastic cover that together with a plastic fan frame and blades adds some elaborate looks to the cooler:

5 zalman cnps10x heatsink
6 zalman cnps10x heatsink

The sides of the heatsink are not covered with anything. Zalman engineers seem to believe that open heatsink sides improve heat dissipation from the plates and lower the heatsink temperature rather than help concentrate the airflow inside the heatsink:

7 zalman cnps10x heatsink

By looking at the top of the cooler, you get a glimpse of a unique regulator that we are going to dwell on a little later. And the bottom view of the cooler will show you your own amazed face reflected in impeccably polished mirror-shining base surface:

8 zalman cnps10x base surface
9 zalman cnps10x regulator

You can remove the top plastic cover by undoing a couple of retention screws:

10 zalman cnps10x plastic cover
11 zalman cnps10x plastic cover

You can also remove the fan from the frame and see that the heatsink is actually not as simple as you might have thought before:

12 zalman cnps10x heatsink

The heatsink plates are of variable height on the side of the entering airflow, which lowers the airflow resistance and reduces the dependence of the cooler efficiency on the fan rotation speed:

13 zalman cnps10x heatpipes

The heatpipes go through the heatsink in a way similar to heatpipes of the Thermalright Ultra-120 eXtreme cooler (by the way, they use the same work in their model names, too):

14 zalman cnps10x heatpipes

As you know, non-linear positioning of the heatpipes inside the heatsink array helps distribute the airflow over the plates more evenly, although no one has ever tried to find out how greatly it actually improves the cooling efficiency. Here I would also like to add that heatpipes are soldered to the cooler base.

Zalman CNPS10X Extreme base is and indisputable example of impeccable evenness…

15 zalman cnps10x base

… as well as remarkable finish quality:

16 zalman cnps10x base

How did they manage to achieve such unique quality? The thermal compound imprints on the glass surface and CPU heat-spreader also turned out ideal. Moreover, the cooler stuck so well to the CPU that we had to wiggle it sideways in order to get it off (that is why the imprint is a little smeared):

17 thermal grease

The heatsink is topped with a 120x120x25 mm fan that is attached to the frame with four self-tapping screws. The fan has seven wide blades made of semi-transparent plastic:

18 fan
19 zalman cnps10x fan

The interesting thing is that it is very easy to remove the fan, however, not so easy to put it back on, because you have to aim the four-pin connector so that the fan contacts sit right against the pins. It will also be pretty hard to replace the default fan if necessary (without removing the default frame and a cover with the regulator).

The fan is built using a dual ball bearing with the promised MTBF of 50,000 hours (seems to be too little for the bearing of this type). According to the marking on the fan rotor, the maximum fan power consumption shouldn’t exceed 3.6 W.

The unique peculiarity of Zalman CNPS10X Extreme is the implementation of the rotation speed control function for its fan. There is a small PWM Mate regulator that is installed into the cooler top plastic cover. The first peculiarity of this regulator is that it can be removed.

In other words, if you decide that you have picked a better spot for it, just slide it out of the cover and stick to the place of choice with a piece of double-sided tape and then use the bundled cable to connect the regulator with the cooler top cover. Very simple and convenient.

The next peculiarity of the fan rotation speed control regulator is the four operational modes it supports:

20 manual rpm control

The thing is that besides a small wheel, there is also a button that allows you to choose one of the three PWM fan modes or enable manual control function. When you set any of the PWM modes, the button closest to the button will light up. Blue color indicates that the rotation speed will vary between 1000 and 1500 RPM with 20-28 dBA of noise, purple color indicates 1000-1950 RPM and 20-35 dBA intervals, and red color indicates 1000-2150 RPM and 20-39 dBA intervals. If at some point the second indicator lights up green, you are in the manual control mode and can change its rotation speed with a little dial wheel in the entire supported range. This is a truly multi-functional regulator 🙂 And seriously speaking, Zalman did a great job on this extremely convenient and functional regulator.

Compatibility and Installation Tips

Besides all contemporary platforms, Zalman CNPS10X Extreme is compatible with the new socket LGA1156 and it is a definite advantage. In fact, it is the first CPU cooler with LGA1156 support that we have tested so far. The cooler can be installed onto AMD processors with the help of an enclosed swing-clip that catches on to the hooks on the standard socket retention frame. As for Intel processors, you have to press Zalman CNPS10X Extreme against them using special steel plate with screws that pushes against the sides of the cooler base:

21 socket am3
Socket 754/939/940/AM2(+)/AM3
22 lga 775
LGA 775/1156/1366

The plate is screwed on to the plastic frame that should be installed onto the board using the enclosed backplate (for LGA775 and LGA1156). For LGA1366 you can use the default reinforcing plate already preinstalled on all the mainboards. In this case the frame is fastened using the enclosed screws with washers:

23 zalman cnps10x installed
24 zalman cnps10x installed

I was surprised to discover that even with the screws tightened to the very end the frame is a little wobbly, i.e. it is not 100% firmly pressed against the PCB textolite. Besides, it seemed very weak and bent during installation. It is actually pretty strange because the cooler is quite heavy and they shouldn’t have neglected the intensity of the pressure against the processor heat-spreader…

When the cooler is installed the distance from the lowest heatsink plate to the PCB is 42 mm. If you turn Zalman CNPS10X Extreme in such a way that its airflow is directed towards the top of the case, then the right side of the cooler will be hanging over the memory nodules:

25 zalman cnps10x pcb
26 zalman cnps10x

And if the modules are equipped with tall heat-spreaders, then you will have to remove modules like that from the DIMM slots closest to the processor socket 9which is exactly what we did in our today’s test session on an AMD platform). Other than that we experienced no problems with the cooler installation and Zalman CNPS10X Extreme fit in very conveniently into our system case:

27 zalman cnps10x heat spreader
28 zalman cnps10x heat spreader

The cooler fan is equipped with blue LEDs:

29 blue led

If some of the peculiarities of the cooler installation remained left out, you can always consult a very well put together installation manual available on the official company web-site (PDF file, 1.64 MB).

Technical Specifications and Pricing

30 zalman cnps10x spec table

Testbed and Methods

We are talking about only one new cooling solution in our today’s review and it will be competing against only two rivals that is why I decided to test the newcomer on two platforms: Intel and AMD. To accomplish this goal we put together two testbeds with the following configurations:

31 intel and amd platform spec table

All tests were performed under Windows Vista Ultimate Edition x86 SP1. We used the following software during our test session:

  • Real Temp 3.20 RC8 – to monitor Intel processor temperature;
  • Everest 5.02.1771b – to monitor AMD processor temperature;
  • Linpack 32-bit with LinX shell version 0.6.0.2 – to create maximum CPU load (two test cycles, 15 Linpack runs in each cycle with 1624 MB RAM capacity for Intel and 1400 MB RAM capacity for AMD);
  • RivaTuner 2.24 – to visually control temperature changes (with RTCore plugin).

So, the complete screenshot during the test session looks as follows:

32 intel platform test session
Intel platform
33 amd platform test session
AMD platform

The stabilization period for the CPU temperature between the two test cycles was about 10 minutes. We took the maximum temperature of the hottest processor core of the four for the results charts. The ambient temperature was checked next to the system case with an electronic thermometer with 0.1 °C precision that allows monitoring the temperature changes over the past 6 hours. During our test session room temperature stayed at 23.5-24 °C.

The noise level of each cooler was measured after 1:00 AM in a closed room about 20 m2 big using CENTER-321 electronic noise meter. Unlike the previous testing approach, starting today we are going to measure the noise levels of each tested cooler outside the system case. In this case the only noise sources in the lab will be the cooler and its fan(s). To measure the noise we set the cooler onto a stand made of polyurethane foam material on top of a desk and the noise meter is always at a 25 cm distance from the cooler.

34 zalman cnps10x noise level

The rotation speed of the coolers fan(s) varied in the entire supported range using the controller that we have by changing the voltage with 0.9-1.2 V increment. The lowest noise reading our noise meter device can register is 29.8 dBA and the subjectively comfortable noise level at this distance is around 35 dBA.

The new Zalman CNPS10X Extreme cooler will be competing against another tower-cooler – ThermoLab BARAM ($50), which also has five copper heatpipes. We equipped this cooling solution with one and two Noiseblocker NB-Multiframe MF12-S3HS fans ($23 x 2) working at 1110 RPM in quiet mode and at 1760 RPM at maximum speed:

35 zalman cnps10x installed
36 zalman cnps10x installed

We also added our ultimate performance reference, Thermalrigth IFX-14 ($79.90), equipped with the same fans:

37 thermalriight ifx 14

Besides, to eliminate the difference in efficiency of the default Zalman fan and Noiseblocker fans, we also tested our today’s hero with the Noiseblocker fan. To install the latter onto Zalman CNPS10X extreme we removed all plastic parts from the cooler:

38 zalman cnps10x
39 zalman cnps10x inside case

The modified cooler is marked as “Mod.” on the performance charts.

Well, let’s check out the results.

Cooling Efficiency

Intel Platform

During this test session we managed to overclock our 45nm quad-core processor with the multiplier set at 21 and “Load-Line Calibration” enabled to 3.97 GHz (+48.8%) using the weakest cooling system of the today’s testing participants. The nominal processor Vcore was increased to 1.3625 V (+13.5%) in the mainboard BIOS:

40 intel core i7 920 cpu z

The memory voltage was at 1.62 V and its frequency was around 1520 MHz (7-7-7-14_1T timings). All other parameters available in the mainboard BIOS and connected with CPU or memory overclocking remained unchanged (set to Auto).

The results turned out pretty interesting:

41 maximum cpu core temperature

Zalman CNPS10X Extreme as is, i.e. without any modifications, can compete against ThermoLab BARAM only at medium and high rotation speed of its default fan. Obviously, the new cooler can’t boast superb efficiency at the minimal fan rotation speed of 1030 RPM, even though it did cope fine with a pretty seriously overclocked processor. And that is in fact quite logical, because thick and dense heatsink plates of the Zalman cooler require more airflow than an extremely optimized BARAM heatsink. At minimal fan rotation speed the default fan doesn’t create sufficient pressure to take the heat off the dense heatsink array quickly enough.

But the most interesting results are, in my opinion, the results of the modified Zalman CNPS10X Extreme. Look, by removing the plastic parts and using one of the best fans we won 4 °C at minimal rotation speed, and at 1760 RPM Zalman CNPS10X Extreme proved 1 °C more efficient than the ThermoLab BARAM! Of course, in this case, the owners of a modified Zalman cooler won’t be able to use the convenient rotation speed controller on the top cover; besides, BARAM can also be equipped with two fans in order to outperform the new competitor. Nevertheless, I wasn’t ready for such a turnaround, because the new Zalman cooler didn’t have anything so innovative about its design that could help us predict its victory over BARAM. Therefore, I decided to perform additional tests on the AMD platform. Here I have to add that Thermalright IFX-14 seems to be absolutely indifferent to what BARAM and CNPS10X Extreme are up to.

AMD Platform

To minimize the dependence of the AMD CPU overclocking on the functionality of not the best mainboard that we had at our disposal, we overclocked our Phenom II X4 processor by raising its multiplier and increasing the bus frequency just a little. As a result, we managed to overclock our processor to 3811 MHz (+27%) without losing stability. The nominal processor Vcore was increased to 1.55 V (+14.8%) in the mainboard BIOS:

42 amd phenom II x4 black edition

The memory voltage remained the same and its frequency was around 1000 MHz (5-5-5-15_2T timings).

Let’s check out the obtained results:

43 maximum cpu core temperature

The main impression that I got from the first test on AMD Phenom II X4 is the low heat dissipation compared to Core i7 we have just discussed. And it is not only because of considerably more modest temperature readings obtained during the test session (the thermal diodes are most likely calibrated differently anyway), or because of lower memory capacity in Linpack (only two 1 GB modules were installed). The thing is that the heatsinks of all three tested coolers warmed up much less than on Intel Core i7, according to my subjective observations. As a result, highly efficient coolers cannot reveal their entire potential on Phenom II X4 that is why the difference between their efficiency test results is lower in this case.

Anyway, Zalman CNPS10X Extreme appears even more efficient on an AMD platform than it was on an Intel one. It loses to BARAM only at minimal rotation speed of its default fan. When tested with the same Noiseblocker fan, the new cooler outperforms its technologically more advanced competitor in both sped modes and even proves more efficient than BARAM with two identical fans. Thermalright IFX-14 is again undefeated, although at the maximum rotation speed of its default fan Zalman CNPS10X Extreme managed to catch up with it.

Acoustic Performance

First of all, let’s see what noise Zalman CNPS10X Extreme generates in the entire supported speed range of its default fan (power consumption data are a nice bonus here):

44 zalman cnps10x noise level and fan input

If the fan rotation speed doesn’t exceed 1070 RPM, the Zalman CNPS10X Extreme cooler can be considered quiet. And at about 1300-1350 RPM it can be called subjectively tolerable. Unfortunately, you can heat slight rustle of the dual ball bearing at low fan rotation speed. In fact, it is similar to the rustling sound from Scythe Minebea Silent IC bearing. I would like to add that the fan startup voltage is 6.3 V and in terms of power consumption, it is hardly an energy-efficient model. We detected no tinkling of the heatsink plates or plastic parts within the entire fan rotation speed range. It is really hard to find any problem here, because the quality of Zalman CNPS10X Extreme assembly is truly high.

Now let’s compare the acoustic performance of all tested coolers in the same fan rotation speed modes as we took for the cooling efficiency tests. The fans were installed exactly the same way as shown on the photo above. They were working for air intake-exhaust on both sides of ThermoLab BARAM heatsink. As for Thermalright IFX-14, one fan was installed on the inner side of the first heatsink array along the direction of the airflow, and another one – on the outer side of the second heatsink array (using rubber bands). Let’s take a look at the comparative diagram now:

45 coolers noise level

The results of our noise measurements performed outside the system case confirmed one very interesting supposition: the shape of the cooler heatsink does in fact affect the level of noise. For example, let’s take the results of Zalman CNPS10X Extreme and ThermoLab BARAM coolers tested with the same fan at 1110 RPM: the noise level of the Zalman solution is lower in this case. At maximum rotation speed of the Noiseblocker fan the situation repeats. Moreover, when we installed two fans onto BARAM cooler, its fan starts jingling at any speed. And while this sound is barely noticeable at the lowest fan speed, then at maximum 1760 RPM it becomes pretty annoying, I have to say. At the same time, the use of two fans on Thermalright IFX-14 causes no jingling of any kind. As for the results of Zalman CNPS10X Extreme with a default fan versus an alternative Noiseblocker fan, the cooler works at even higher rotation speed with the latter, which is, in fact, quite expected.

Conclusion

The new Zalman cooler made the most favorable impression on us. After CNPS9900 LED it is a really serious step forward for Zalman, even though CNPS10X Extreme turned out just another super-cooler to join several dozens of others. Although, I don’t think the use of the words “just another” would be appropriate in this case, because the new cooling solution from Zalman is equipped with such unique features as a multifunctional fan rotation speed controller and LGA1156 support. And while the latter feature may soon be available from a few other cooler manufacturers, PWM Mate will most likely remain Zalman’s unique distinguishing advantage. Also, do not forget about a “dark horse” – Zalman ZM-STG2 thermal compound, which we are going to check out shortly within our next thermal compounds roundup.

As for the drawbacks, we can certainly point out weak retention for the Intel platform. Thin plastic retention frame doesn’t provide enough hold for the Zalman CNPS10X Extreme base pressed against the CPU heat-spreader. We hope that Zalman engineers will improve this retention solution soon. Besides, despite unusually low recommended retail price of only $59.90, it would be really nice if they also offered a “lite” modification of the same cooler without the plastic parts and maybe even without the rotation speed controller and pricy thermal paste but at a much more affordable price. But the biggest drawback in my opinion, is that this great cooler appeared not two years ago, but only today when there are quite a few serious solutions in the market already.

Nevertheless, I am sure that a “strong and handsome” cooler will find its fans, even though it will not be a king…

About The Author

XbitLabs Team

We are a team of enthusiasts thriving to provide you with helpful advice on buying tech.

Subscribe
Notify of
guest
0 Comments
Inline Feedbacks
View all comments