by Sergey Lepilov
07/25/2008 | 07:12 PM
Summer is definitely not on the black list for overclocking fans. Even though you can’t tweak sunshine, tan and vacation time, but these earthly pleasures are surely not alien to overclockers. However, we can’t deny that the summer heat that set in these days does have certain negative effect on overclocked system components including CPUs.
Some overclockers even have to lower their speeds for the sake of higher system stability, according to a few posts in the forums. And for everyone else higher air temperature is just another great opportunity to check out the stability of their overclocked systems in more severe conditions.
The two air-cooling systems for processors we are going to discuss today will be the tools to help you fight these conditions. Please allow me to introduce the CPU coolers from XIGMATEK and APACK.
The box of the new XIGMATEK Achilles S1284 cooler boasts extremely original design. The cooler is encapsulated in a clear plastic box with large cardboard inserts inside. One of them is decorated with an ancient shield image:
This is part of the Trojan War theme implied in the cooler’s name: Achilles is the undefeated mythic hero whose only weak spot was his… heel. In fact, it turned out pretty symbolic, as you will see later on :)
You can see the cooler through the clear plastic package without even removing it from the box. It is interesting that the cooler heatsink is packaged upside down, this way drawing the potential customer’s attention to its major distinguishing feature – heatpipe direct touch technology in the cooler base. There are detailed technical specifications of the cooler on the reverse side of the box alongside with the list of compatible processor types.
At the bottom of the package there is a small plastic bag with accessories including the following items:
The cooler heatsink is a typical tower heatsink consisting of an aluminum plate array sitting on four copper heatpipes:
There are 46 plates in the heatsink array, each about ~0.35mm thin, which is a little thicker than the heatsinks of this type would normally have (~0.25mm). The plates are spaced out 2mm from one another.
XIGMATEK Achilles S1284 measures 120 x 85 x 159mm and weighs 660g (with the fan).
The sides of the heatsink are partially covered with the plate edges bent downwards:
I have to say that they are made in a very interesting manner, as angled ledges. You can clearly see it on the top view shot:
As a rule, covered heatsink sides encourage more efficient use of the air flow from the cooler fan and improve the overall cooler efficiency.
On the same photograph you can see that the heatpipes are spread out in the heatsink array with a shift. It allows not only making the heatsink smaller, but also distributing the heat over the heatsink plates more evenly than in case of lined up heatpipes:
As you may have already noticed there are four heatpipes, each 8mm in diameter. The heatpipes are covered with a thin layer of nickel alloy:
The cooler base is covered with protective film that has to be removed before installation:
The heatpipes in the base sit in an aluminum bar with slits. The gaps between the slits measure 2mm:
Of course, no one thought of polishing off the heatpipes in the base of the cooler, but the base is very even that is much more important than mirror-shining finish. The photograph below shows the thermal compound imprint on the cooler base from an LGA 775 CPU:
As you see, the CPU actually contacts only the two heatpipes in the center of the base and about 40% of the two heatpipes on the sides. Moreover, a part of the processor heat-spreader has no contact with these heatpipes at all. As we have already pointed out they are shifted, so it is impossible to have heatpipes contacting the entire processor heat-spreader. Maybe this approach together with 2-mm gaps between the heatpipes is indeed the “Achilles’ heel”. Our tests will answer this question later, and now let’s continue studying the cooler design and features.
XIGMATEK Achilles S1284 is equipped with a very beautiful 120-mm fan with seven semi-transparent orange blades:
The fan rotation speed can be adjusted automatically in the interval from ~800RPM to ~1500RPM using pulse-width modulation algorithm (four-pin cable and connector):
Moreover, there are four white LEDs in four corners of the fan. The manufacturer claims that this fan produces 20.6dBA of noise, but there is no mention of the operational mode corresponding to this value. I dare assume that it is most likely the minimal noise reading.
The fan is attached to the heatsink with four silicon spindles that fit right in-between the heatsink plates:
This ensures better acoustic performance (than in case the fan is pressed against the heatsink plates), because the spindles absorb fan vibrations. They are pretty robust but at the same time soft enough, so don’t be afraid to rip them apart during fan installation.
XIGMATEK Achilles S1284 cooler looks like this with the fan attached:
No second fan can be installed on the other side of the heatsink.
The cooler is compatible with all contemporary socket types. In order to install it onto the board, you don’t need to remove it from the system case. If you have an AMD K8 or K10 processors, you should use the bundled swing-clip that catches on to the hooks on the standard plastic frame and locks the cooler in place with a tag:
The distance from the lower heatsink plate to the top of the mainboard PCB is about 40mm, which is enough room for all sorts of components in the area around the processor socket.
If you wish to install XIGMATEK Achilles S1284 onto an LGA 775 mainboard, you have to attach two retention brackets with plastic locks to the cooler base using bundled retention screws:
They use no backplate in this case, so the mainboard PCB may bend a little bit with the cooler installed. It is not critical, but could have been better if they had a backplate, because as we all know coolers with heatpipe direct touch technology are very sensitive to proper contact with the CPU heat-spreader.
The XIGMATEK cooler looks very stylish inside the system case:
By the way, the cooler should be installed with the heatpipes across the inscription on the CPU heat-spreader, because coolers with heatpipe direct touch work their best when positioned like that (the improvement is about 2-4ºC depending on the CPU heat dissipation). I have already pointed this fact out to you in my previous reviews.
The cooler looks even more charming with the white LED highlighting, especially at night:
The MSRP for XIGMATEK Achilles S1284 is set at $44.99, which is a very decent price for a cooling solution of this type.
The second cooler we are going to discuss today was made by APACK, which is famous for its originally designed cooling solutions. ZEROtherm ZEN ZF120 cooler comes in a small silverfish box made of thick cardboard. There is a round window cut out in the front of the box. There you can see part of the heatsink and the fan:
There are a few pictures of the cooler installed onto a board on the reverse side of the box alongside with its dimensions. The detailed specifications of ZEROtherm ZEN ZF120 are listed on one of the box sides.
The accessories are hidden at the top of the box:
Among them there are the following items:
There is nothing revolutionary about the heatsink design. It is a tower heatsink sitting on four copper heatpipes 6mm in diameter that go through a copper base plate:
There are 56 aluminum plates, each 0.25~0.3mm thick forming a heatsink array. The gaps between the plates are smaller than those on Achilles: 1.5mm. The cooler measured 121 x 61 x 156mm and weighs 670g (practically the same as XIGMATEK Achilles S1284 we have just discussed).
The main feature distinguishing ZEROtherm ZEN ZF120 from other similar solutions is the so called honey-comb structure, when the central part of the plates is pushed in, while the sides are lifted up. According to the manufacturer, this creates better airflow turbulence and improves cooling efficiency.
At the same time, unlike XIGMATEK cooler, the sides of the ZEROtherm ZEN ZF120 are not covered with anything, so we can assume that part of the airflow escaping through the sides of the heatsink will be wasted.
It is interesting that even when we taped over the sides of ZEROtherm ZEN ZF120 heatsink, there was no change in the cooling efficiency and the air hardly escaped through the cooler sides. Maybe it was the central honeycomb structure that somehow concentrated the airflow in the middle of the cooler reducing the losses almost to null. I would also like to add that the heat dissipating surface of the cooler heatsink measures 6,827sq.cm, which is 79sq.cm larger than by the flagship product of the ZEROtherm cooler lineup – Nirvana NV120 Premium.
The cooler base is covered with protective plastic film with a warning that it has to be removed before installation:
The heatpipes sit in special grooves cut out in the copper base plate and are soldered to the base:
The copper base in its thinnest spot beneath the heatpipes is a little less than 3mm.
The base finish quality is superb. It is just a tiny bit away from mirror-shine:
The copper plate is ideally even: the thermal compound imprint on the glass surface as well as on the processor heat-spreader was impeccable.
ZEROtherm ZEN ZF120 is equipped with a beautiful completely transparent fan with seven blades. The fan measures 120 x 120 x 25mm:
The fan allows PWM rotation speed adjustment. The supported RPM rate is from ~1100 to ~1800RPM with 59.5CFM airflow and 19.5~31.4dBA of noise. The manufacturer doesn’t reveal the type of bearing they used for the fan and its MTBF. However, since there were no stickers of any kind on the fan rotor, we couldn’t find out this information on our own.
Before installing the fan onto the heatsink, you have to stick two shock-absorbing strips to it (the strips have one sticky side):
They serve the same purpose as the silicon strips of the XIGMATEK cooler: reducing vibrations and improving the acoustic performance.
The fan itself is fastened with two wire clips that catch on to the holes in the cooler heatsink at the top and bottom of it:
The clips hold the fan very firmly in place, unlike the clips of the Thermalright Ultra-120 eXtreme cooler.
Even with the fan, ZEROtherm ZEN ZF120 doesn’t look bulky and weighs only 690g, which is not that much according to today’s standards:
The cooler retention mechanism remained the same as by previously reviewed coolers from this manufacturer. First you attach the corresponding bracket to the cooler base with four included screws and then it either locks on the standard plastic frame of the Socket 754/939/940/AM2(+) or has to be screwed to the backplate on LGA 775 platforms:
The cooler is pressed very firmly against the CPU, especially in the latter case when screws are involved. This as well as impeccable contact between the cooler base and the processor heat-spreader ensure impressive cooling efficiency of ZEROtherm products.
ZEROtherm ZEN ZF120 looks like this inside a system case:
For coolers with a heat-conducting base it doesn’t matter what way they are facing. Therefore, we installed our ZEROtherm ZEN ZF120 so that the airflow would be directed towards the case rear panel equipped with a 120-mm fan ousting the air outside the case. I only have to add that the distance from the cooler lower heatsink plate to the mainboard PCB is 42mm.
The technical specifications and recommended pricing of the enw coolers will be compared against ZEROtherm Nirvana NV120 Premium:
The today’s coolers and their competitors were tested in two modes: in an open testbed when the mainboard sits horizontally on the desk and the cooler is installed vertically, and in a closed testbed with the mainboard in vertical position.
Our testbed was identical for all coolers and featured the following configuration:
All tests were performed under Windows XP Professional Edition SP2. SpeedFan 4.34 was used to monitor the temperature of the CPU, reading it directly from the CPU core sensor:
The mainboard’s automatic fan speed management feature was disabled for the time of the tests in the mainboard BIOS. The CPU thermal throttling was controlled with the RightMark CPU Clock Utility version 2.35.0:
The CPU was heated up with OCCT (OverClock Checking Tool) version 2.0.0a in a 23-minute test with maximum CPU utilization, during which the system remained idle in the first 1 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 took about half the time. Despite the stabilization period, the result of the second test cycle was usually 0.5-1°C higher. 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).
The ambient temperature was checked next to the system case with an electronic thermometer that allows monitoring the temperature changes over the past 6 hours. During our test session room temperatures varied between 27.0 ~ 27.5°C. It is used as a staring point on the temperature 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 ZEROtherm Nirvana NV120 Premium fan rotated at 1830RPM, which equals the rotation speed of the ZEROtherm ZEN FZ120 fan. Both generated identical noise. According to our measurements, both ZEROtherm coolers appeared 0.5~0.7dBA louder than the XIGMATEK cooler, which is actually not a critical difference. Since the fans of both new cooling solutions support PWM rotation speed control, we didn’t test them in quiet mode at minimal fan rotation speeds.
Besides the above listed coolers we have also included the results for Thermalright SI-128 SE that is the best cooler of all we have tested so far. Therefore, we are using it as a reference point. This cooler was tested in one operational mode with the XIGMATECH Achilles S1284 fan working at ~1530RPM:
Now let’s move on to the actual test results.
Using the “weakest” cooling system with a quiet fan we managed to overclock our quad-core 45nm processor to 3.8GHz inside the system case. The processor Vcore was increased to 1.525V in the mainboard BIOS:
The monitoring utilities reported a little lower voltage setting than the one in the mainboard BIOS: 1.48~1.5V:
Click to enlarge
The test results of our today’s heroes and their rivals are given on the diagram below:
As we have expected, the first newcomer, XIGMATEK Achilles S1284, didn’t prove very efficient against the participating competitors. Moreover, its performance was very poor when tested inside a system case: it fell 15ºC behind the leader! In an open testbed where Thermalright SI-128 SE has no advantage of a 120-mm fan on the case side panel, Achilles S1284 catches up a little, but the remaining 9ºC gap is still pretty significant.
The second new cooler, ZEROtherm ZEN FZ120 performs much more confident in our today’s test session. No, it didn’t outperform the flagship product of ZEROtherm’s lineup, Nirvana NV120 Premium, even at identical fan rotation speeds (I have to remind you that the latter has some extra speed in reserve: up to ~2700RPM, without any negative effect on the acoustics). However, since ZEN FZ120 is cheaper, we hardly had the right to expect otherwise. Nevertheless, ZEN FZ120 and Nirvana NV120 Premium turned out equally efficient in an open testbed and fell just a tiny bit behind Thermalright SI-128 SE with our overclocked processor. In case of higher CPU overclocking and further Vcore increase, the advantage of Thermalright super-cooler increases a little more. However, with the scorching hot weather outside I decided to skip maximum CPU overclocking this time for safety reasons :)
First about XIGMATEK Achilles S1284. Relatively inexpensive and stylish cooler employing pretty young Heatpipe Direct Touch technology will not be able to satisfy demanding overclocking fans’ needs on contemporary CPUs. But what determined the insufficient efficiency of this cooler with HDT technology? I believe there were several factors that contributed to the result. I think you already know about a 45nm CPU that doesn’t suit well for HDT. Add here too big of a gap between heatpipes and aluminum base plate instead of copper. Moreover, don’t forget the incomplete contact between the processor heat-spreader and the shifted heatpipes in the cooler base (“Achilles’ heel”?), absence of LGA 775 backplate and weak cooler retention. So, looks like XIGMATEK engineers have a lot to work on to ensure that Achilles S1284 didn’t end up like its Trojan namesake. By the way, while we were working on this review, XIGAMTEK send us a new cooler revision with a few improvements, so stay tuned for an update.
The second participant of our today’s test session showed its real best. High cooling efficiency, almost the same as that of a more expensive Nirvana NV120 Premium; moderate price; simple and reliable retention mechanism and easy installation procedure; moderate noise under peak CPU utilization and very quiet operation in regular mode. Is there anything about ZEN FZ120 that could have been better? I believe, it could be nice to have an option to install a second fan and increase the heatsink plate density from 1.5mm to 1mm between the plates. With more plates, the effective cooling surface area would also increase. And finally, ZEROtherm could switch to 8mm heatpipes that are currently used in the very best CPU coolers. Anyway, there is another great CPU cooler out there now: ZEROtherm ZEN FZ120.