by Sergey Lepilov
05/17/2009 | 09:01 AM
Some of you may be surprised to learn that super-coolers make the smallest share of all CPU coolers purchased by users today, but this is really so. There are several reasons for that. First, these coolers are primarily designed for overclockers, and in fact, relatively small group of users out there overclock processors. There are even fewer users among them who try to squeeze all the juices out of their processors with air-cooling alone. Second, super-coolers are expensive and pretty large in size, so you have to really think everything through before buying one. Third, there are mainstream coolers in the today’s market that cost half or even one third the price of the super-coolers and are much smaller in size. The interesting thing, however, is that these products lose not so much to the super-solutions in terms of cooling efficiency as you might think they would. However, very often you have to put up with lower maximum CPU frequency during overclocking and lower Vcore.
Today we are going to talk about four new cooling solutions from the mainstream segment that includes coolers ranging from $25 to $40 in price. Therefore, our today’s heroes - Cooler Master Hyper N520, Scythe Katana 3, GELID Silent Spirit and Scythe REEVEN - all fit this profile perfectly well. All of them are quite compact in size and easy to install. In our today’s article we will not only compare their functionality, cooling efficiency and acoustics, but will also see how far behind they are compared to the best air-cooler in CPU temperature as well as maximum overclocked frequency.
The first solution we will be talking about today ships in a pretty large box for a cooler of this size. It is designed in Cooler Master’s traditional colors:
There is a lot of information on the box that gives you a good idea of the cooler inside. The cooler is bundled with the following accessories:
Cooler Master Hyper N520 looks very interesting to say the least of it. It consists of a medium sized heatsink with the fans on both sides of it:
The fans are shifted from one another and work for air intake and exhaust:
The cooler weighs 688g and measures 122.4 x 102.5 x 141 mm:
Cooler Master Hyper N520 heatsink is covered with an aluminum shroud that holds two 92-mm fans:
The manufacturer most likely decided to shift the fans away from one another trying to eliminate the so-called “dead zones”, when some heatsink plates do not get cooled with the fan airflow (these zones are most often located right beneath the fan rotor).
Besides that, you should also check out five 6mm heatpipes that are positioned chequer-wise inside the heatsink. As we know, this heatpipe structure inside the heatsink improves heat distribution over the heatsink body:
If we remove the shroud with the fans, we can take a closer look at the heatsink:
It consists of 48 aluminum plates, each 0.45mm thick and spaced out at 1.8mm. You can notice that each heatsink plate has five slits in it. I suspect that they have been made to create turbulent airflow from the fans and increase the heat dissipation efficiency. The calculated effective heatsink surface measures 5,045sq.cm, which is very good for a mainstream cooler.
The heatpipes lie in special grooves cut out in the copper base plate:
We didn’t notice any traces of soldering, however, we know for a fact that Cooler Master has always used soldering technique instead of thermal glue to ensure proper contact between the base plate and cooler heatpipes. Looks like they have simply done it very carefully and neatly.
The base plate surface is very even:
It is also very well finished, although it is not quite mirror-like (not a must, of course, but a very attractive feature):
However, as I have already said, it is not an issue. Most importantly, the surface is even, which we can see from the thermal compound imprint on Cooler Master Hyper N520 base:
The heatsink holds two 92 x 92 x 25mm fans (Cooler Master Cooler Master A9225-18CB-3BN-L1 model):
The fans rotate with constant speed of 1800RPM. They create 3.8CFM airflow and should produce no more than 19dBA of noise. The claimed static pressure is at 3.24mmH2O (it must be the total pressure from both fans mentioned in the official specs). Rifle bearings used in these fans should last for 70,000 hours or about 7 years.
Cooler Master Hyper N520 is compatible with all contemporary platforms. Just use the appropriate retention kit with the corresponding backplate:
To protect the PCB surface from scratching, you should use rubber washers together with the enclosed screws:
After that the cooler is tightened with screw-nuts through the backplate or plastic washers, if no backplate can be used for any reason.
This is what Cooler Master Hyper N520 looks like when it is installed into a system case:
The installation manual recommends installing the cooler into the system case so that the airflow from the fans could be directed towards the back of the case. However, our experiments revealed that the cooler performed best if turned by 90°:
In this case the CPU temperature under maximum workload was steadily 3-4°C lower than in the first case. The distance from the mainboard PCB to the lowest heatsink plate of the Cooler Master Hyper N520 cooler is 36.6mm, so the cooler will not interfere with any electronic components around the CPU socket.
The recommended price of the Cooler Master Hyper N520 is set at $39.90.
The second new cooler we are going to test today is Scythe Katana 3. It will be available in a small box that is filled with all sorts of information, following Scythe’s good old habit:
The features of the new cooler are listed on its package in two different languages – English and Japanese, while the specifications are available in six languages:
The cooler is bundled with retention kits of three different types, SilMORE thermal compound and installation instructions:
As you may have already guessed, Scythe Katana 3 is a successor to the well-known Katana and Katana 2 coolers. Unlike the previous models, Katana 3 uses three copper heatpipes instead of two. Although when you take a first look at the picture, you may think there are six of them, and not just three:
The cooler has become a little bigger, too. Now it measures 94 x 108 x 143mm and weighs 495g. What remained the same, however, is the inclined positioning of the heatsink and heatpipes relative to the base plate:
Now this feature has its own name and is called S.P.S. (Slant Pipe Structure). However, its purpose hasn’t become any different: it ensures better cooling of the components around the processor socket.
The cooler heatsink consists of 47 aluminum plates, each 0.35mm thick that are spaced out at 2mm from one another. The calculated effective heatsink surface equals 4,292sq.cm.
The second patented technology implemented in Scythe Katana 3 is called F.P.S. (Fast-Phase Structure). Its brief description indicates that it is in fact a small pin-heatsink installed above the heatpipes in the cooler base that helps the primary heatsink dissipate heat from the heatpipes.
The heatpipes are flattened at the base and soldered to it. They form a curve in the contact area:
We have already seen a very similar implementation of heatpipes contact at the base in Scythe Kama Angle cooler.
The base plate is 1.5mm thick. Its surface is covered with a thin layer of nickel alloy covered with a protective film warning you about removing it before cooler installation.
The base surface is impeccably even and is polished off so well, that the little toy was startled at its own impression:
However, the thermal compound imprint on the CPU heat-spreader could have been a little better: it turned out not completely even. However, it happened not because of some cooler base imperfections, but because of the push-pin retention clips. Unfortunately, when the cooler is fastened with plastic clips like that, you can’t always ensure even pressure between the CPU and the cooler base.
Scythe Katana 3 is equipped with a 92-mm fan (Scythe SY9225SL 12M-P model). Its rotation speed varies from 300 to 2500 RPM using PWM control method.
According to the official specifications, the airflow created by this fan also varies depending on the rotation speed and can be anywhere from 6.7 to 55.55 CFM generating between 7.2 and 31.07 dBA of noise. There is no info on the slide bearing MTBF and static pressure reading.
Scythe resorts to global and at the same time very simple approach to ensuring compatibility of their cooling solutions with all contemporary platforms. They use V.T.M.S. (Versatile Tool-Free Multiplatform System) retentions that should be inserted into special gaps in the lower heatsink:
There is the same type of retention used for LGA775 and LGA1366 mainboards, however, you have to shift the clips more to the end of the retention panel for the latter case:
The installation instructions for Katana 3 do not mention any preferred position on the CPU inside the system case that is why during our test session we decided to turn it with the heatpipes facing up:
In this case Scythe Katana 3 performed 2°C better than in case it was installed with the airflow from the fan directed towards the back of the system case. This difference is fairly small, but, nevertheless, it exists. Since the new cooler is pretty compact in size, it doesn’t interfere with any components around the CPU socket.
The recommended price of the new cooler is set at $32, which makes Scythe Katana 3 a perfect representative of the mid price range.
GELID Company is a new name in the cooling solutions market that is why its solutions are not as widely spread yet as the products from Cooler Master or Scythe. Nevertheless, we have already tested GELID GC-1 thermal interface in our previous roundup, and very soon will introduce to you their new 120-mm fans. Today, however, we would like you to meet GELID’s first and so far the only cooler called Silent Spirit.
There is a small cut out window on the front of the cooler package that reveals part of the device inside:
There is absolutely no free space anywhere on the package: the entire box is covered with information about the cooler:
Silent Spirit is shipped with a retention kit for AMD platform, a small sticker and brief installation manual:
The retention for LGA775 platforms is already preinstalled onto the cooler base. The GELID GC-1 thermal interface is also pre-applied onto the base surface. As for the LGA1366 retention kit, it ships in a small separate box…
…and includes two steel plates with plastic push-pins and a set of screws for them:
GELID Silent Spirit seemed to be a very interesting cooler. Even though its structure seems pretty simple at first glance, it combines almost all known technologies used in CPU cooling solutions these days. Well, here it is, a small (108 x 100.5 x 125 mm) and light-weight (370g) cooler:
It looks similar to so-called top-coolers; however, its primary heatsink array is slightly elevated above the horizontal axis. Four copper heatpipes 6mm in diameter that come out of the copper base plate hold 44 trapezoid-shaped aluminum heatsink plates:
Each plate of 0.4mm thick and the distance between them is 2mm. the calculated effective heatsink surface is 3.160sq.cm.
I am sure you have already noticed round hollows stamped into the heatsink plates. According to the manufacturer, it should increase airflow turbulence and improve heat dissipation from the heatsink plates. Another peculiarity of GELID Silent Spirit heatsink is the different height of the plates:
However, unlike Noctua coolers that also use heatsink plates of different height, GELID Silent Spirit doesn’t alternate them one after another, but has them arranged in groups of two or three. To be more exact, at first there are two identical plates, then three plates that are 6mm shorter, then another two taller plates, and so on and so forth. This lowers airflow resistance and should ensure high cooling efficiency of this solution even at low fan rotation speed. There are three tall plates on both outer sides of the heatsink.
The copper base plate has special grooves cut out for the heatpipes that are soldered to it ensuring more secure contact than flattening of the heatpipes and the use of thermal glue.
There is a small aluminum heatsink installed above the heatpipes. According to the developers, it helps unload the heat from the heatpipes in the base and also serves as additional support for the cooler retention.
This is the bottom view of GELID Silent Spirit cooler:
Look at the layer of thermal interface applied onto the cooler base:
According to the manufacturer, it is GELID GC-1 – pretty efficient thermal paste, which we have already tested before. Nevertheless, there seems to be too much of it, because even surfaces should have a minimally thin layer of thermal interface on them. By the way, the cooler base is indeed very even and gets a good B+ for finish quality.
Now a few words about the fan. The fan is fastened in a plastic frame that catches onto the grooves on the sides of the heatsink with plastic hooks:
It is a 92 x 25mm fan with PWM adjustable rotation speed that ranges from 900 to 2400 RPM generating from 10 to 25.8dBA of noise. It creates 45.8 CFM maximum airflow. The fluid dynamic bearing of the fan should last for 50,000 hours of non-stop operation.
The distinguishing feature of this seven-blade fan is four silicon spindles it is installed on:
They absorb parasitic vibrations and prevent them from transferring onto the heatsink. Moreover, GELID Silent Spirit developers stress that according to their research, there should be 5-5.5 mm between the fan and the heatsink for most optimal performance, which is exactly what they did on their Silent Spirit cooler.
You won’t need to remove the mainboard from the system case to install the cooler onto the board. For AMD platforms you use two retention panels that catch on to the hooks on the standard plastic retention frame, and for LGA775/1366 platforms you use the enclosed push-pin retentions:
There is very little info about the preferable cooler position inside the system case. They only mention that if the cooler is installed so that the heatsink is angled towards the back of the case, the warm air will be quickly removed from around the CPU by the rear case fan. If you case also has a fan at the top, then you can install the cooler like this:
The MSRP for GELID Silent Spirit is set at $35.
Finally, the fourth testing participant in our today’s session is the solution that also comes from the Japanese Scythe Company. It is called REEVEN and ships in a small vertical box.
The information on the box tells you everything about this cooler except its cooling efficiency and actual noise level:
Inside the box you find the cooler, of course, as well as the following accessories:
We were a little stunned that this new cooler is still bundled with the retention kit for the long outdated Socket 478 platform, where no one will ever really use the new REEVEN cooler. However, there was no retention kit for the new LGA1366. The cooler is made in Taiwan.
The design of Scythe REEVEN is fairly simple, but has two unique features about it. First of all I have to say that it is a tower cooler measuring 96 x 92.5 x 136.3 mm and weighing 425g. It consists of 39 aluminum heatsink plates and three copper heatpipes 6mm in diameter:
The plates are relatively soft, they are only 0.5mm thick. The gap between them measures 2mm, just like by the other coolers we have discussed earlier today. The calculated effective heatsink surface area makes 3,882sq.cm.
The outer edges of the heatsink plates are turned down and perforated where the airflow exits the cooler. Moreover, these heatsink plates alternate with shorter plates without any perforation. This is the first peculiarity of the Scythe REEVEN cooler. This solution is likely intended to create a turbulence zone on leaving the heatsink that also sends the airflow down to the electronic components around the CPU socket and the heatsinks on them.
The second distinguishing feature of the REEVEN cooler is what Scythe calls Heatpipe Direct Contact. We have already come across Scythe’s first prototypes employing this technology on two tradeshows, but in reality the first cooler from Scythe to actually implement this is REEVEN. The difference between this cooler and similar competitors’ solutions is the absence of any aluminum or copper inserts between the heatpipes in the base, because all heatpipes there are combined into a single unit:
As a result, even though there are gaps between the heatpipes in the base, the probability of dead zones forming during the heat transfer from the CPU to the cooler is minimal. The heatpipes are slightly spread out away from one another inside the heatsink:
The cooler base is even and finished well enough for a solution using heatpipes direct contact technology:
Judging by the thermal interface imprint from the Core i7 processor heat-spreader on the cooler base we can see that heatpipes cover about 4/7 of the processor heat-spreader, while the rest contacts only with the aluminum retention panel. Nevertheless, we feel that it is not a severe issue of the new cooling solution, because the actual die size of this CPU is only 20% of the size of its heat-spreader, and as you can see, heatpipes cover well over that area.
There is a 92-mm fan on the heatsink attached with two wire clips.
Although this fan has a different marking than Scythe Katana 3 fan, it has absolutely identical specs: PWM 0-2500 RPM, 55.55 CFM, 31.07 dBA).
Scythe REEVEN is compatible with Socket 478, LGA775, Socket 754/939/940/AM2(+)/AM3 platforms and you won’t need to remove these mainboards from the system case to install the cooler. Unfortunately, as we have already mentioned above, the LGA1366 retention kit is not included with this cooler; however, Scythe reps assured us that this issue will be resolved shortly. We had to use the Core i7 retention kit from Scythe Kabuto cooler: we had to cut it up in two halves and bend the retention hoops a little more to ensure more secure contact with the CPU heat-spreader:
As a result, we managed to successfully install this cooler onto our testbed:
I have to say that we didn’t reveal any dependence of the REEVEN cooling efficiency on the way it is installed onto the CPU, although we had expected to find some, judging by our experience with other coolers of similar structure. During our test session, REEVEN was installed as follows:
In conclusion I would only like to add that the recommended price of this new cooler is set at only $24. It is actually the cheapest cooler in our today’s test session.
All tests were performed inside a closed system case. Our testbed was identical for all coolers throughout the test session and featured the following configuration:
All tests were performed under Windows Vista Ultimate Edition x86 SP1. We used the following software during our test session:
So, the complete screenshot during the test session looks as follows:
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 or open testbed 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 1:00AM in a closed room about 20sq.m big using CENTER-321 electronic noise meter. The measurements were taken at 1m and 3m distance from the closed system case. During the acoustics tests all five 120-mm case fans were slowed down to ~520 RPM. In this mode the background noise from the system case measured at 1m distance didn’t exceed ~33.3 dBA. When the system was completely powered off, our noise meter detected 30.8 dBA (the lowest on the charts is 30 dBA). The subjectively comfortable noise level is around 34.5~35 dBA.
Now a few words about the competitors. First, as we have already promised you before, we are going to check how these new mainstream cooling solutions compare against our reference super-cooler - Thermalright IFX-14, which is the most efficient CPU air cooler today. I am sure you understand that we can’t compare any of our today’s testing participants with the super-cooler from Thermalright in price, weight or size; however, we are not going to perform any direct comparison today. During our test session we used a seven-blade 120-mm fan from XIGMATEK Dark Knight cooler installed between the heatsink arrays of IFX-14.
Second, we decided to include XIGMATEK Nepartak cooler into our today’s test session, as it also belongs to the mainstream price range ($25). Although it is equipped with a PWM controlled fan, its rotation speed was locked at two values: 1740RPM for the quiet mode and 2810RPM for the maximum performance mode. The same is true for all other testing participants. The only difference is that their fan rotation speeds were in fact different for these modes.
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 3.84 GHz (+43.8%) using the weakest cooling system of the today’s testing participants. The nominal processor Vcore was increased to 1.2875V in the mainboard BIOS (+7.3%). The memory voltage was at 1.64V and its frequency was around 1500MHz (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 of our cooling efficiency tests are sorted according to generated noise (moderate noise/maximum rotation speed) and in order of increasing cooling efficiency (lowering of the peak CPU temperature). Let’s take a look:
The leader in the first group – tests performed in the operational modes with moderate noise – XIGMATEK Nepartak is a convincing leader among mainstream cooling solutions, we have already concluded in one of our previous articles that this cooler offers great performance for the given price range. Scythe REEVEN and GELID Silent Spirit are just a little bit behind the leader, and Scythe Katana 3 is another 2°C back. The least efficient here appears Cooler Master Hyper N520. However, it is important to understand that it is not working in its nominal operational mode: its fans have been forced to slow down. The difference between the best and the least effective cooler in this test is 8°C. The super-cooler is 10°C+ ahead of all the testing participants.
The second group represents the results obtained at maximum rotation speeds of the coolers fans and the situation here is very different. First, Cooler Master Hyper N520 that used to be behind everyone is now the second best, following closely behind XIGMATEK Nepartak. And the outsider spot has now been taken by Scythe Katana 3. Second, GELID Silent Spirit turned out 1°C more efficient than Scythe REEVEN: even though their fans work at the same speeds, the former is undoubtedly more technologically advanced. In this mode the performance difference between the best and the least efficient cooler reduces to 6°C that is why availability and price should not be the only factors determining your choice. It is also extremely important to consider acoustic performance, which we are going to do next.
The coolers are lined up from the quietest to the noisiest:
Everything is quite logical here. The only thing we would like to point out is relatively low noise from the GELID Silent Spirit cooler.
I believe that it is impossible to single out a definite leader among the four new cooling solutions discussed in this review. All coolers differ just slightly in efficiency as well as in their acoustic parameters. From the pricing prospective, Scythe REEVEN seems to be the most attractive solution at a recommended retail price point of only $24, but unfortunately, it comes with no LGA1366 retention yet. The next best choice would be GELID Silent Spirit, being a quiet and efficient cooler in its class, although it costs more than REEVEN. The remaining two participants, Scythe Katana 3 and Cooler Master Hyper N520 are also pretty close in their cooling efficiency.
XIGMATEK Nepartak cooler looks very impressive against the background of our today’s contestants: it remains a definite efficiency leader in both groups. Considering that it has very low MSRP of only $25, we decided to use it as a reference point for mainstream coolers testing. As for the efficiency difference between the mainstream cooling solutions and the super-cooler team represented by Thermalright IFX-14, I doubt any commentary is necessary on my part. You can see clearly how many degrees you can buy for paying at least twice as much. And speaking in MHz, we can say that this extra cash will buy you 200MHz of CPU speed. Note that the dependence of overclocking potential on cooling efficiency may differ both ways on other platforms, CPUs and even processor steppings.