by Sergey Lepilov
03/11/2008 | 08:46 PM
Don’t you think that the potential of CPU air cooling solutions has been completely exhausted by now? We keep testing new CPU air-coolers, many of which demonstrate great cooling efficiency even for significantly overclocked quad-core CPUs on Kentsfield core. However, we haven’t seen any serious breakthrough in this field for quite some time already. Although we test new air coolers on a regular basis, all these solutions prove only efficient for their price range, but none managed to outperform the today’s undefeated leader – Thermalright Ultra-120 eXtreme yet. In other words, efficient processor cooling has become considerably more affordable and much easier to arrange these days, but that’s about it. Some of you may be more than satisfied with that, however, since we are working with new cooling solutions every day in our lab, we are hungry for a real breakthrough in this field. We would really want to see the processor peak temperatures go down and their overclocking potential go up more or less greatly, because fort the past year and a half all we dealt with were 0.5-1ºC differences or even failure to successfully compete with the recognized cooling efficiency leaders.
So where shall the well-known cooler makers direct their efforts? Shall they continue to increase the heatpipes diameter, make their coolers even larger or pack them with two or even three fans now? Shall they use more precious metals for their heatsinks, offer all sorts of modding add-ons or release limited edition solutions? How can they increase their products’ efficiency and attract potential customers? It is a pretty hard question to answer now. As we have seen recently, new Intel processors manufactured with 45nm process do not require expensive super-coolers for successful overclocking. As of today, overclocked 45nm processors do not necessarily require super-coolers.
Nevertheless, manufacturers continue announcing and releasing new cooling solutions that aim for the leading positions in this market. And today we are going to talk about two newcomers from XIGMATEK and Scythe: HDT-D1284 and ZIPANG:
New cooling systems have similar design principle (the fan is set on top of the heatsink), but are still very different from one another. That is why it will be especially interesting to take a look at two new products side by side and see how they compare against the well-known super-coolers in terms of their functionality, cooling efficiency and noise level. Let’s get started.
The new XIGMATEK cooler comes in a small gray box made of unusually thin cardboard:
There is a cut-out window in the front of the box that shows part of the cooler fan and the protective grid on it. The front side of the box also lists the key features of the new solution and bears a small photo of the cooler. The reverse side of the box has a table with detailed specifications of the cooler and states that it is made in Taiwan.
You may have noticed that the box looks a little bit beaten by the postal services. However, there is no need to be concerned with cooler safety, because it inside is packed in a thick and sturdy foamed polyurethane casing that protects it and the accessories perfectly fine from transportation damages:
Accessories are placed in a small section at the bottom of the box. XIGMATEK HDT-D1284 comes bundled with the following items:
The cooler design doesn’t boast any innovative approach. However, nevertheless, XIGMATEK engineers managed to combine several successful solutions used by other cooler makers in their HDT-D1284:
So, this 667g cooler boasts the so-called top-design when the airflow from the fan is directed to the mainboard PCB surface. The cooler is based on four copper heatpipes, each 8mm in diameter. The heatpipes pierce 60 (!) aluminum plates of trapezoid shape:
Aluminum heatsink plates have stamped grooves at the bottom that should create additional turbulence for the airflow and thus ensure that hot air will not stagnate in the heatsink plate array. We have already seen the same solution by OCZ Vendetta cooler.
The heatpipes feature an elbow curve half-way to the heatsink:
This curve may be intended to ensure better compatibility of the XIGMATEK cooling solution with the mainboards and provide more free space beneath it. If XIGMATEK made these heatpipes straight, the tall chipset or RAM heatsinks could become a significant obstacle hindering cooler installation. AT the same time, it is hard to tell how this additional curve will affect the gas and liquid movement as well as overall cooler efficiency. However, I believe that additional resistance will no pass completely unnoticed for the overall heat dissipating efficiency.
One of the key features of the new XIGMATEK HDT-D1284 cooler is the H.D.T. technology used on the contact surface between the cooler base and the processor heat-spreader – Heatpipe Direct Touch:
Four heatpipes are part of the cooler base and are pressed into the aluminum plate. There is thermal compound that ensures proper contact between the heatpipes and the base plate:
It is hard to tell how sturdy this contact is. Nevertheless, the contact between the heatpipes and the aluminum base plate remained absolutely reliable throughout the entire test session when the cooler was installed and dismounted multiple times.
The heatsink plate array is topped with a 120x25mm fan. Its blades are covered with protective wire grid:
The fan sits in the heatsink grooves on tuber compensators that should absorb parasitic vibrations and reduce the level of generated noise:
The fan is very easy to remove, so it will be no problem to replace it if necessary. The fan features a four-pin power connector and supports PWM rotation speed management function.
So, the fan rotation speed may vary from ~800RPM to ~1500RPM with maximum 27.2dBA of noise and 56.CFM airflow. The Rifle bearing of the fan is guaranteed to run 40,000 hours.
The protective plastic film on top of the cooler base doesn’t serve any real protective function:
…because the quality of the cooler base finish leaves much to be desired:
The installation procedure is extremely simple. On AMD K8 platforms XIGMATEK HDT-D1284 is fastened with a metal “sling”-clip that can be inserted between the heatpipes or perpendicular to them:
There are four slits cut into the aluminum cooler base specifically for that. This way XIGMATEK engineers made sure that their cooler is free from a drawback typical of many cooling solutions out there: limited positioning choices on Socket 754/939/940/AM2 platforms. Now XIGMATEK HDT-D1284 may be installed facing in any direction to achieve maximum convenience and cooling efficiency.
In case the cooler needs to be installed onto an LGA775 mainboard, the positioning choice is not an issue to begin with. That is why all you need to do is screw on the appropriate retention to the cooler base:
And then you simply install the cooler onto the mainboard by pushing the plastic clips into the retention holes around the processor socket. Although, I would strongly recommend doing it only after the mainboard has been removed from the system case. Otherwise, when the mainboard is inside the case, you will not be able to push the spindles into the holes really hard, as they are underneath the heatsink. Therefore it is pretty tricky to reach them, and in our case it was simply impossible to do at all:
You can easily notice that I installed the cooler with the heatpipe ends facing upwards. Despite pretty large cooler size of 145 x 150 x 134mm, the heatsink as well as the curved heatpipes didn’t hit against any components in the around-the-socket area. Moreover, even if we rotate XIGMATEK HDT-D1284 90 degrees counterclockwise, the heatpipes still won’t prevent you from installing memory modules with tall heat-spreaders, not to mention regular memory modules:
It is interesting that when XIGMATEK HDT-D1284 is installed with its heatpipes facing particularly this way as in the picture above, it didn’t transfer the heat from the CPU as efficiently as we had expected reporting the exact same monitoring graph every single time and reporting an error after a minute into the test:
We did check the contact between the cooler base and the CPU heat-spreader multiple times using our thermal compound imprint method:
However, it proved that the problem lies specifically with the heatpipes positioning, because once we moved the cooler with its heatpipes facing upwards, the thermals got back to normal and all tests were a pass without any problems. To tell the truth, this is pretty strange because we haven’t seen anything like that during our previous tests of similarly designed cooling solutions. Besides, the heatpipe performance doesn’t really depend on its positioning. In case of XIGMATEK HDT-D1284, we have double-checked and confirmed the direct dependence of the cooler efficiency on the heatpipes positioning. Maybe the elbow curve or larger heatpipes diameter (8mm instead of the standard 6mm) or even the heatpipes direct touch technology affects the results… Of course, it would be interesting to see if the same dependency occurs by other samples of the same cooler, but unfortunately, we do not have the chance to check it out yet, because we only have got one sample at hand.
In conclusion to our discussion of the new XIGMATEK HDT-D1284 cooler we have to add that its recommended retail price is set at $48.
The second new solution to be reviewed today was created by the Japanese Scythe Company that has been announcing quite a few new processor cooling solutions lately. ZIPANG launch has just been announced, when they rolled out two new models: Scythe Shuriken and Orochi. We are going to offer you reviews of these two solutions in the nearest future and in the meanwhile let’s talk about an extremely interesting cooler called Scythe ZIPANG.
A not very big box of yellow-greenish color made of thick sturdy cardboard. The front of the box has a cooler photo and a few small icons indicate the key features of this cooling solution:
The same features are described in a little more detail on the sides and back of the box. Inside the box is separated in two sections, the smaller containing bundled accessories:
Scythe ZIPANG comes bundled with the following items:
The cooler boasts impressive size of 145 x 148 x 112mm and weighs 815g:
At first glance, the cooler design doesn’t strike you with anything particularly innovative:
The cooler boasts the same top orientation of the heatsink array with a fan, six copper heatpipes 6mm in diameter that go through the cooler base, and 68 thin aluminum plates of the heatsink array sitting on the heatpipes. The only unusual thing is a 140mm fan on top of the heatsink that makes this new cooler stand out among other similar looking solutions. However, this is just the first impression. In reality everything is not as simple as it might seem.
The thing is that Scythe ZIPANG for the first time uses UPHC technology – Uneven Parallel Heatpipe Construction – that should improve heat transfer efficiency by reducing the losses. In reality it looks like this:
Heatpipes pierce the cooler base and exit at the top of the heatsink. Their long ends pierce aluminum heatsink array and their short ends go right beneath it, holding it at the bottom. Besides, these heatpipes are curved not exactly beneath the top loop, but with a slight shift:
The cooler heatsink sits on the heatpipes on exactly in the middle, but also with a slight shift to one side:
This solution may have been intended to ensure that the airflow from the fan cools down the memory heat-spreaders or chipset cooler, or even the heatsink on mainboard power components of the voltage regulator circuitry. It will depend on the way the user installs the cooler.
The heatsink is cooled with a 139 x 139 x 25mm fan with seven fan blades:
According to the technical specifications, Scythe ZIPANG fan rotates at 1000RPM generating 21dBA of noise and creating 51.82CFM airflow.
The fan is marked as DFS132512L and is made in China. It seems to be a very quiet fan, from our subjective impression.
The fan is attached to the heatsink with two wire clips. If you remove them you will see that there is a small groove along the sides of the heatsink intended for a 120x120x25mm fan that can also be installed on top of Scythe ZIPANG heatsink:
In this case it is important to point out that although the clips for a 120mm fan are not included with the cooler accessories, you can also use the default clips from the 140mm fan. Although they will bend oddly, when installed into the 120mm retention grooves. I would also like to mention that there are no shock-absorbing pads or stickers on the heatsink as well as on the fan, which could have reduced the level of generated noise even more by absorbing the fan vibrations. However, any overclocker can make such rubber pads on his/her own, so that shouldn’t be a problem.
There is a pretty large aluminum heatsink in the lower part of the cooler right above the cooler base and heatpipes:
It helps the heatpipes a little because it is also cooled with the airflow from the fan on top.
Unlike the most efficient Scythe cooler – the Infinity (Mugen), the heatpipes in the base of Scythe ZIPANG are practically not flattened at all:
Moreover, Scythe doesn’t ever use the same approach to heatpipes contact with the copper cooler base as Zalman, Thermalright and even Gigabyte. Namely, they do not place the heatpipes into specifically designed grooves in the base plate. No matter how you look at it, but just soldering (or even sticking) the round heatpipe to the flat surface of the base plate is indeed very simple and least labor-intensive approach, but it definitely is less efficient than placing the heatpipes into the grooves. It’s a pity Scythe loses a few degrees here…
The cooler base is covered with protective film that should be removed before installation:
The quality of the base finish and its evenness checked with a thermal compound imprint on the glass surface and the processor heat-spreader are ideal and will definitely satisfy even the most demanding overclocking fan:
By the way, the manufacturer claims that the cooler base is silver coated. We didn’t double-check this statement during our test session.
Besides two contemporary platforms – Socket 754/939/940/AM2 and LGA 775, Scythe ZIPANG may also be installed onto mainboards with the pretty outdated Socket 478. That is why those of you who still own this platform and haven’t yet replaced the default boxed cooler with anything more advance have one more choice option to consider. They say you may not need to remove the mainboard from the system case to install Scythe ZIPANG onto any of the above mentioned platforms. However, it is not quite true. Theoretically, the installation is extremely simple, just fasten the appropriate retention brackets to the cooler base with four screws…
… and then install it on top of the CPU covered with a thin layer of thermal compound. I don’t know about AMD K8 processors, but in case of LGA 775 platform I could not complete the installation with the mainboard inside the system case. The cooler size and height above the mainboard PCB wouldn’t let me reach the plastic clips and push them in until locked. Besides, it really makes sense to make sure that the cooler weighing 815 grams is installed properly with all clips locked securely in place, which cannot be done when the mainboard is in, unless you have a gastrofibroscope available :)
There is nothing in the manual about the cooler positioning preferences that is why we first installed it so that we could also ensure additional cooling of the copper heatsink on top of the chipset North Bridge, which heats up a lot during our tests:
In other words, the heatpipes were positioned horizontally. However, after a few preliminary test runs, we discovered that when the cooler sits with the long ends of the heatpipes facing upwards, the processor temperature under peak workload was always 2ºC lower. It is hard to tell what really affected the end result in this case: was it really the heatpipes positioning (we didn’t see the same effect on an open testbed) or additional cooling of the mainboard voltage regulator heatsink:
But the fact is undeniable: if the heatpipes in the cooler heatsink array (not the ones beneath it) are aimed with their ends upwards, the processor temperature under maximum workload is 2ºC lower than in case of their horizontal orientation. By the way, the same photo shows how a 120mm fan can be installed and that the wire clip on one end is not even fully fastened, but the fan holds on very well.
In conclusion to our discussion of the new Scythe ZIPANG cooler I have to say that the recommended retail price of this solution is set at $58.80, which is quite comparable to what other super-coolers are priced at these days.
The technical specs and recommended price of the today’s two cooling solutions and their competitors also participating in this test session are given in the table below:
We tested the cooling efficiency of new XIGMATEK and Scythe processor coolers and their competitors in two modes: in an open testbed when the mainboard sits horizontally on the desk and the coolers are installed vertically, and in a closed testbed with the mainboard in vertical position
We put together the following testbed for our experiments:
Using the weakest cooling system of our today’s testing participants we managed to overclock our quad-core processor to 4GHz with the Vcore increased to 1.6V in the mainboard BIOS. The monitoring utility reported the core voltage setting a little bit lower than what was set in the mainboard BIOS: around 1.575~1.6V. The system memory was working at 1066MHz efficient frequency with 5-5-5-16_2T timings and 2.05V voltage.
All tests were performed under Windows XP Professional Edition SP2. SpeedFan 4.34 Beta 40 was used to monitor the temperature of the CPU, reading it directly from the CPU core sensor. Its readings matched those of the Core Temp 0.96.1 utility. The mainboard’s automatic fan speed management system was disabled for the time of the tests in the mainboard BIOS. The CPU thermal throttling was controlled with the new RightMark CPU Clock Utility version 2.35.0 that now supports Intel Core 2 Extreme QX9650 processor.
The CPU was heated up with OCCT (OverClock Checking Tool) version 1.1.1b in a 23-minute test with maximum CPU utilization, during which the system remained idle in the first and last 4 minutes of the test. I performed at least two cycles of tests and waited for approximately 20 minutes for the temperature inside the system case to stabilize during each test cycle. The stabilization period in an open testbed with the mainboard in horizontal and coolers in a vertical position took about half the time. The maximum temperature of the hottest CPU core of the four in the two test cycles was considered the final result (if the difference was no bigger than 1°C – otherwise the test was performed at least once again). Despite the stabilization period, the result of the second test cycle was usually 0.5-1°C higher.
The ambient temperature was checked with an electronic thermometer that allows monitoring the temperature changes over the past 6 hours. During our test session room temperatures stabilized at around 25°C. It is used as a staring point on the diagrams. Note that the fan rotation speeds as shown in the diagrams are the average readings reported by SpeedFan, and not the official claimed fan specifications.
The noise level of each cooler was measured according to our traditional method described in the previous articles with the help of an electronic noise meter – CENTER-321. The subjectively comfortable noise level was considered 34.5dBA and is marked with a dotted line in the diagram. The ambient noise from the system case without the CPU cooler didn’t exceed 33.2dBA when measured at 1m distance.
The two new coolers from XIGMATEK and Scythe will be competing against Thermalright Ultra-120 eXtreme super-cooler and the best Scythe’s air cooler called Infinity (Mugen). Both competitors were equipped with the same Scythe Minebea 120x120x25mm fan rotating at ~1940RPM (according to the monitoring utilities) and at ~1190RPM in its quiet mode. Scythe ZIPANG was also additionally tested with the same fan.
The coolers are lined up according to their launch time in the market. The results are given on the diagram below:
The first unexpected surprise came from XIGMATEK cooler that lost to all other testing participants. Moreover, it fell up to 10ºC behind the leaders, which is more than surprising, because the design of HDT-D1284 is very promising especially thanks to heatpipe direct-touch approach. As you may remember, we have already tested cooling solutions with heatpipe direct-touch technology and we have observed exactly the same effect. It means that even though they suit for cooling 45nm quad-core Intel processors on Yorkfield core, they prove considerably less efficient with them than with the older 65nm Kentsfield based processors.
The second surprise, this time a pleasant one, came from a new Scythe solution – the ZIPANG cooler. The Japanese cooler proved extremely efficient inside a closed system case: it provided 6ºC better cooling efficiency than its brother, the Scythe Infinity cooler, and even outperformed the Thermalright Ultra-120 eXtreme by 3ºC! Of course, we shouldn’t underestimate the contribution from the side panel fan in the system case that worked better for the Scythe ZIPANG and XIGMATEK HDT-D1284 than the tower-shaped coolers. Besides, Thermalright Ultra-120 eXtreme regained its leading positions in an open testbed. Nevertheless, the results of our today’s test session show that overclockers have definitely got another new super-cooler at their disposal.
Unfortunately, the overclocking potential of the new quad-core processor was limited at 4050MHz with 1.6V Vcore even with the today’s best cooling system. We didn’t feel like raising the voltage any more for psychological reasons and because of the info on 45nm CPUs degradation as a result of overclocking with extremely high voltage settings. That is why today we are not offering you the maximum CPU overclocking results for each cooler tested.
And what about XIGMATEK HDT-D1284 on 64nm processors? We didn’t have a quad-core Kentsfield available to us at the time of the tests that is why we performed additional tests on a dual-core Intel Core 2 Duo E6750 with 2.66GHz nominal frequency that was overclocked to 3.86GHz with 1.6V Vcore setting. The obtained results looked as follows:
XIGMATEK HDT-D1284 has fully rehabilitated itself after pretty unimpressive performance on a Yorkfield based processor. So, we dare suppose with pretty high probability that XIGMATEK cooling solution should be just a little less efficient than Thermalright Ultra-120 eXtreme on hotter 65nm quad-core CPUs on Kentsfield core.
We measured the level of noise generated by our today’s testing participants from a 3cm, 1m and 3m distance. The results are given on the diagram below:
Don’t be surprised that there are no results for Scythe Infinity and Thermalright Ultra-120 eXtreme coolers on the diagram above. Since they used the same 120mm fan as Scythe ZIPANG, the acoustic difference didn’t exceed 0.1~0.2dBA at all times that is why we decided not to overload the diagram with identical data. As for the acoustic performance of the new cooling solutions reviewed today, the outcome was quite predictable. Both newcomers proved extremely quiet, even XIGMATEK HDT-D1284 at the maximum fan rotation speed of ~1430RPM cannot be considered noisy, even though it crossed the subjectively comfortable borderline a little bit.
Winding up the discussion I started in the beginning of this review, I can say that we didn’t witness any significant breakthrough in terms of reducing the CPU temperature with the new cooling solutions. The obtained results prove that we haven’t hit another qualitative milestone just yet, even though there are two new solutions to be added to the super-coolers list. Let’s sum up the obtained results for each of them separately.
XIGMATEK HDT-D1284 finally proved to be a very successful solution, however, we have to stress that it is only true for 65nm processors. Not very high cooling efficiency with 45nm processor on Yorkfield core against the background of today’s other testing participants may result from a smaller contact spot between the processor die and the CPU heat-spreader (Yorkfield’s die size is smaller than Kentsfield’s). Other than that the cooler is an excellent solution: lightweight, universal and convenient to use, not noisy at all, allowing easy replacement of the default fan with a more powerful one, with the lowest retail price of all today’s testing participants ($48).
Scythe ZIPANG, on the contrary, proved a terrific cooler from all standpoints. Very high cooling efficiency, low level of generated noise, ability to work with a 120mm fan instead of a 140mm one, excellent cooling of the around-the-socket area and even memory modules, universal design – all these advantages will be extremely attractive not only to overclocking fans but also to regular users who care about proper cooling and quiet system operation. As for the drawbacks, I have to mention that it is very hard to install if the mainboard is inside the system case. And some overclockers who never have enough may wish there were a more powerful fan with a rotation speed controller. As for the cooler price of $58.80, it is absolutely adequate and comparable with the price of ZIPANG’s direct competitors.