INTRODUCTION
NZXT was established in 2004 and bring to market several cases, coolers, case accessories and power supplies. The products are designed in the USA by a very small team that punches a lot above their weight, and manufacturing is done in Asia as is the case with most companies in this field. My primary interest in the NZXT peaked as a result of their Kraken G10 GPU bracket announced last year. I had an AMD Radeon R9 290 and wanted to make a “Cooling Hawaii” roundup. NZXT, especially Mitchell, was gracious enough to send a G10 bracket as well as Kraken x41 to use with it. Now anyone familiar with AIOs knows they usually come with pre-attached TIM on the cold plate, and that poses an issue for me who does multiple runs to be statistically accurate. The applied TIM is also meant to be used on a CPU IHS, rather than a bare GPU die. So I decided to test out the x41 as a CPU cooler first before using it in conjunction with the G10 as a GPU cooler.
UNBOXING AND OVERVIEW
The cooler comes in a nice, big box that immediately tells you what you have and points out the 6 year warranty on it! It is also a very universal cooler catering to all contemporary Intel and AMD sockets.
On the back are some specs listed out as well as a reminder to go download CAM- something we will get to soon. What is missing here, and something I would like to see added, is a reminder to go check out the very nice
installation guide on the NZXT website. Again, this is something we will get to real soon.
Not exactly what I was expecting to see when I opened up the box, but a pull later we get to see the contents properly:
Here is the installation guide that comes with the cooler. While it does point you to what needs to be done, I can see first time and novice builders needing assistance. I always believe that a good installation guide/manual should be included as a hard copy with any component that requires installation before you get the chance to view anything on a monitor. The way NZXT did it, they assume you have a second screen to see the animated installation guide on their website.
Along with the installation guide we find the cooler itself as well as the included single 140mm fan and accessory pouch.
The fan is their new FX V2 140mm PWM Performance one. The blades (7 of them) have an ivory color to them (as opposed to the white shown on the NZXT website) and go well with the black radiator. Going by the blade design and hub size, I would not have guessed this was a high static pressure fan but we will see how it performs soon.
What’s this? A massive 0.6 A current draw rating? This better be a monster at airflow if it is rated to consume a max of 7.2 W (on a 12 V rail) during operation. Most 140mm consumer fans are rated at less than 0.2 A. Interestingly, the specs on the product page list a motor current of 0.325 A. I can’t be sure if that is for the pump or the fan given the terminology used in the list is all over the place. We will see for ourselves later on if the fan does indeed draw 7.2 W because, if it does, motherboard fan headers will be challenged to run the cooler the way NZXT intends.
There are 8 vibration dampening pads (2 per corner).
I have seen this plastic sleeving used on several OEM fans before and all I can say is this does the job and is not an optional extra cost. One can easily cut the heatshrink and sleeving if a custom sleeve job is in the plans.
The fan is PWM controlled via a standard 4 pin header. Way to go, NZXT! Now let’s take a look at the accessories provided:
You get the Intel backplate for sockets 115x, the AMD mounting bracket for the AIO, 4 screws each for AMD sockets, and for Intel socket 115x, 2011/2011-3. You also get 8 washers, 4 UNC 6-32 6mm long screws and 8 UNC 6-32 30mm long screws for the fan/radiator installation.
COOLER OVERVIEW
The radiator, the cables and the waterblock/pump unit are nicely protected in case of any mishaps during shipping.
Ah, that’s better!
The pump/waterblock unit comes with the Intel mounting bracket pre-installed as well as some TIM. The cold plate is copper and is circular with screws all countersunk to be level with the plane of the cold plate and not interfere with mounting. The contact area is big enough for even the huge Intel LGA 2011/2011-3 CPUs.
The water tubing used is ~388mm long and should give provide multiple mounting options with most cases. If used in conjunction with a GPU cooling bracket like the G10, then this extra length of tubing compared to other AIOs on the market may well turn out to be invaluable.
The radiator itself comes in measured at 141 x 172 x 36mm. Build quality is decent compared to custom loop radiators. There are 14 tube channels set in a single row, each of which is 2 mm thick. The radiator is a standard dual pass, U-flow type as far as coolant flow goes. I understand this is an all aluminum radiator based on an Asetek design.
The fins are also aluminum, serpentine type coming in 35 microns thick. The fin density was measured at 20-22 FPI. This is the first AIO CLC I have tested so knowing this is thicker than most AIOs and yet having a 19 FPI fin density makes me apprehensive about the fan airflow needed if this trend continues with other AIOs. With great airflow also comes great tendency to trap dust, especially if you try to have all radiator fans on intake wherever possible.
Now on to the cables:
There are inputs to accept two 4 pin PWM devices. This is where you plug in the included fan after installation onto the radiator, as well as a second fan if you purchased one intending to go push-pull. Here’s the catch though: this FX V2 fan is not yet listed on the NZXT store (or anywhere else for that matter) for a separate purchase. You can of course use a different fan if you wish, but the jury on the impact of using different fans on either side of a radiator is still out there.
There is a 3 pin header for power that runs the pump as well as the fans connected above. Here’s where my confusion begins. A 3 pin header- recommended by NZXT to be plugged into a 4 pin PWM header, that does not have the RPM tach sensor on it. Yes, you read it right. When I hooked this to a fan controller such as my Aquaero and controlled it via power (voltage control), I was able to slow down the pump and fans accordingly while having no RPM readout. This is NOT RECOMMENDED one bit by NZXT.
The tach sensor and RPM control is all done via CAM using the USB 2.0 header cable:
I will go into more detail on this soon but suffice to say I have not fully understood how and why this was done the way it is and this position is not something I like to be in- especially as a reviewer/tester. Phone calls with NZXT support did not help much here, although they were very patient with me (major props to Jake and Eric here!). I wish I was able to disassemble the unit to take a good look at the CPU waterblock/pump unit but given this is a CLC and I only have one that may be needed in the future, I am unable to do so unfortunately.
INSTALLATION
Before I go any further, I must point out that I am testing on 3 Intel CPUs across 2 sockets. I apologize on behalf of any AMD users but this is mainly a component restriction here- I have no AMD CPUs or motherboards to test on. Note also that the pictures below are for demo only.
Let’s go back to this picture:
The posts on the left are short M4 and these are intended for the Intel socket 2011/2011-3. As with all coolers, this is the easiest socket to install on. You begin by simply screwing in these 4 posts into the 4 standoffs in the socket like so:
Next, simply route the waterblock/pump unit (which has the Intel bracket on it) over the posts in the outer set of holes before using the 4 thumb nuts to screw it in place:
Tighten down the thumb nuts two at a time (diagonal opposites) alternating between each set after a couple of turns. Continue till they are fully tightened. Since this was also when the one and only application of the pre-attached TIM was done, here’s a look at the mount and TIM spread:
Not bad at all. For my tests, I used Gelid GC Extreme and for socket 2011/2011-3, I like to use a thin line down the middle.
Once this is done, plug in the 3 pin header to the CPU header (or any standard 4 pin PWM header), the fan to the daisychained input and the USB 2.0 cable to the USB 2.0 header on the motherboard:
Now on to socket 115x, specifically LGA 1150 shown on an m-ITX motherboard:
Use the provided backplate and have the notches towards the center for this socket. It then fits into the standoffs in the motherboard perfectly. In fact, backplate orientation did not matter at all for either of my LGA 1150 motherboards:
Once this is done, flip over the motherboard carefully and screw in the other 4 posts into the backplate. Both ends are identical so worry not about which end goes where:
Follow the same steps as before:
Be sure to plug in the 3-pin power header as the sticker says before peeling it off.
One last thing- in case the Intel bracket comes loose or if you wish to change it out for the AMD/G10 bracket:
There are 2 pieces here- the bracket itself and also a ring that holds it in place. First up, slide the bracket of choice up towards the tubing making sure that the “teeth” on the bracket are facing downwards:
Next slide the ring into place, there are several orientations where it will work and so it is almost fool proof:
Now push the bracket down on the ring so the teeth catch on each other. There are 4 sets in total so make sure to push down on the bracket appropriately:
A LOOK AT CAM
CAM is NZXT’s venture into the “intelligent” hardware monitor tool. The installer is ~75 MB in size. One funny thing I noticed in the EULA was this:
Yeah, if you are 13 or younger then you supposedly are not allowed to use it. Also, if there was any doubt over the OEM for the cooler:
Once the installation is complete, you are prompted to create a profile so you can then save your settings as well as be able to monitor online or via the CAM app on an iDevice (Android app expected Q1, 2015). You have the option to skip this too thankfully.
Once this is done, you are then prompted into a tutorial which again can be skipped. I recommend actually taking the few seconds and going through the few stages each time a tutorial prompt comes up if you intend on using the software. This below is “Basic” mode for example:
You will have noticed that the tutorial displays the actual rig specs themselves. Once this is done, if the x41 cooler is hooked up properly then you get a notification (and an achievement!) of the cooler being recognized by CAM. A “Kraken” mode then unlocks prompting another small tutorial:
Now we can finally take a proper look at CAM as it relates to the cooler. There are 4 modes of control over the cooler- manual, custom (curve), silent and performance. In manual mode, you set a power level from 25% to 100% in steps of 5% and that’s it:
Notice the part where the liquid temperature is higher than the CPU? CAM does need some work, and it’s response is sluggish at times. But it does function as it promises. You can also monitor all primary components (CPU, RAM, GPU, storage etc) with time. The best part- it does not take a lot of system resource. Aesthetics can be a personal thing also but I don’t mind the functional build as it is now.
In custom mode, you set a custom curve as per component temperature. In this case, it would be the CPU. Do keep in mind that in all cases, you control both the pump and the fan(s) simultaneously. I wish I could better explain the mode of control here and if I do understand it in the future then I will update this article accordingly.
Silent and performance modes are present curves:
I would like to see the values on the X-Y plot at the left being shown, especially in custom mode. At this point you are guessing what each step is and there is no reason for this to be the case. Either way, this is the only method to read out pump RPM. But there is still no way to control/measure the fans independently here. You can of course control the fans separately via a proper PWM header elsewhere but this is a single step control which no doubt many people would prefer.
CAM also has integrated into it their NZXT Hue to be able to customize the face plate on the pump/waterblock unit. By default it is green, but you have RGB control (via slider only) as well as set colors to choose from. You can also set a specific color in case of an emergency warning such as the CPU temperature exceeding a set value. You also have blink or breathe mode for the colors. Bear in mind that the set colors aren’t very accurate. For example, I chose red and I got this:
By using the sliders for R, G and B however, I did get a proper red:
Another thing- notice the part where the fan RPM is 0 in there? Well, it is not so. The fan was definitely up and running properly but CAM did not recognize it nor was there a warning that the fan RPM was zero. If it was indeed true, then the CPU would have overheated in minutes under load. While I remain optimistic about things like NZXT CAM and Corsair Link, I would rather it not at all be released as a consumer ready software till bugs are ironed out.
TESTING
First, the pump. As mentioned in the previous page, using manual mode and jotting down the pump speed was the only way to get the pump’s response curve. Pump noise was measured in an anechoic chamber of size 5′ x 8′ with a sound probe held 6″ away from the pump to measure the sound volume in dBA accordingly.
Those are some very impressive curves (lack of, rather). Assuming the pump RPM numbers are accurate, then the very linear response is great to see. The pump is also very quiet- especially closer to the lower end of the power chart. If this is what future AIO pumps will be like, then bring them on!
Now for the fan. An Aquacomputer Aquero 6 XT was used to monitor fan speed and control it via PWM from 100-35% where the duty cycle ended. Linear airflow was measured using an Extech 45158 Thermo-Anemometer 6″ away from the fan furthest away such that it measured the airflow in feet per minute through the radiator. Fan noise was measured in an anechoic chamber of size 5′ x 8′ with a sound probe held 6″ away from the furthest fan to measure the sound volume in dBA accordingly.
This is one of the fastest consumer 140mm fans out there. Given what we know of the radiator, you definitely needed a fast fan to provide enough airflow through those fins. It also means that the fan comes in louder than the pump and thus is the bottleneck as far as cooler noise goes. At this point, I would generally see if another fan might be better but the 140mm fans I have all slower than this. Sure, I could have used two of them but then again I would much rather be able to buy one of this very fan instead and run them slower/quieter instead. NZXT, please make this happen! Oh, for those thinking about how a motherboard fan header (that is usually rated to provide 12 W of power) can power a pump that is rated at 7.5 W and 2 fans that are each rated at 7.2 W- rest assured. This fan came in at 0.17 A max (2 W). So even with 2 of these, you are at 11.5 W draw from a motherboard header. When I contacted NZXT about this, I was told that each fan would be different. This wasn’t exactly what I wanted to hear and I have never seen any fan so far off from advertised specs (RPM specs and measured agreed with each other nicely too). So take this for what you will. Another option, as suggested previously, is to run the fans separately off another PWM header instead.
Now to thermals.To find out how this performs as a cooler, I did thermal tests on 3 systems:
8 core: Intel i7 5960x (4.2 GHz at 1.2 Vcore) with dual channel DDR4 memory
(a) Power level set at 50% (Test 1)
(b) Power level set at 100% (Test 2)
4 core: Intel i7 4770k (4.5 GHz at 1.25 Vcore) with dual channel DDR3 memory
(a) Power level set at 50% (Test 3)
(b) Power level set at 100% (Test 4)
2 core: Intel Pentium G3258 (4.5 Ghz at 1.2 Vcore) with dual channel DDR3 memory
(a) Power level set at 50% (Test 5)
(b) Power level set at 100% (Test 6)
Each test was done in a climate controlled hotbox at 25 ºC ambient but the results will still be reported in terms of a delta T (component – ambient) in ºC. The ambient temperature should be nearly the same throughout the box with a +/- 0.25 ºC variability at most but still the ambient temperature sensor was held in front of the intake fan as should be the case with CPU coolers in my opinion. Each test/scenario was repeated three times and the average value taken to account of possible issues with mounting and TIM. Speaking of TIM, Gelid GC-Extreme was used as explained earlier and cure time was accounted for by measuring values after steady state was achieved post burning in with an Intel XTU run. Realtemp was used to measure the core temperatures.
At 100% power, the cooler noise was 53 dBA while it was closer to 40 dBA. The temperatures are great for a single 140mm AIO but I truly feel the pump has been let down by the radiator/fan combo here. Of course there are limitations when working with an OEM but if NZXT can push for a ~14-15 FPI radiator with splitter type, unlouvered fins and replace this one fan with 2 of their slower 1000 RPM FN V2 fans instead then I can only see this being even closer to top end air coolers while being quieter too. Speaking of other coolers:
In order to compare the cooler with others, this is the scenario I chose: Intel i7 4770k at 4.5 Ghz, 1.25 Vcore, and everything set to full speed. I am not going to have any common components fans. My reasoning for this is that this is a test of the cooler unit and not just the heatsink or radiator. I will, however, use Gelid GC-Extreme for all the comparison tests because some coolers (such as this very one) don’t come with a tube of TIM for multiple usages and so a stock TIM there is moot anyway.
My thoughts carry onwards here. I am eager to see how the Kraken x61 would fare here, or perhaps the Kraken x42 with a more optimized radiator/fan combo. But for now, the Kraken x41 is another AIO that comes close but not close enough. Bear in mind that the other coolers being compared against are bigger dual radiator/heatsink tower type. The advantage of a hotbox is that the temperature inside can be controlled easily, but this is not necessarily the case in a case. Especially with air coolers such as the PH-TC14PE, case airflow matters a lot and so if you have not planned it out properly then stagnant hot air in the CPU area will result in liquid cooling options such as this NZXT x41 outperforming even the best of the air coolers.
CONCLUSION
The NZXT Kraken x41 comes in at $110 in the USA from the
NZXT store or via Newegg/Amazon. It is certainly an interesting device wherein the pump is quite novel for a CLC in that you have some control over speed and the CAM software at this point is enough of a positive to list me as such. It does trail behind the Swiftech H220-X and Phanteks PH-TC14PE as a CPU cooler while also being louder than than the Phanteks cooler which is also less expensive. But as seen before, the fan is the culprit there and something that the end user can replace as seen fit. Keep in mind that the H220-X and PH-TC14PE are dual rad/heatsink type coolers so perhaps this is not a very just comparison. Perhaps, seeing how the Kraken x61 is only $20 more, one might be tempted to go with it if space is not an issue. Perhaps I will have the oppportunity to test out the Kraken x61 and see if this concept is what AIOs should be heading towards. But till then, please let me know what you think- especially the part about the cooler comparison. Thanks again for reading!
post edited by vsg28 - 2015/01/12 15:28:41