by Sergey Lepilov
12/01/2008 | 07:51 PM
Looks like the launch of Intel Core i7 platform hasn’t yet inspired cooling solution manufacturers. No increased activity has been detected yet in the cooling solutions market and no brand new solutions for the new processor and complete platform are coming out yet. Instead, many well-known brands are offering additional retention kits for the existing cooler models either for a small price or even for free, like Noctua. However, it is pretty logical, because the cost of making a new retention kit is incomparable with the cost of designing a new cooling solution from scratch.
I believe that most importantly they should make sure that the coolers selling these days come bundled with new retention kits, because cooling solutions may often outlive one or even two processors. Unfortunately, three coolers from our today’s article do not have an LGA 1366 retention kit with them (at least so far). Nevertheless, they may be extremely interesting to our regular readers, because they have been significantly improved compared with the previous models. I am talking about Xigmatek HDT-SD964, Xigmatek HDT-S1284 and OCZ Gladiator Max, which we are going to discuss today.
We have already discussed Xigmatek HDT-SD963 cooler, and now XIGMATEK Co., Ltd released an enhanced modification of a relatively compact cooler. The box is pretty small, made of thin cardboard. It has all the information about this cooling solution printed on it, including description of employed technologies and technical specifications:
There is a polyurethane foam casing inside the box that protects the cooler from physical damage during transportation. There is a plastic bag with accessories in its upper part:
Let me list them all one by one from left to right and from top to bottom:
Xigmatek HDT-SD964 is made in Taiwan.
Xigmatek usually has the fan packed separately from the heatsink. So, let’s start with the latter:
The heatsink of Xigmatek HDT-SD964 cooler is of very simple design. It is built on four copper heatpipes 6mm in diameter that form part of the cooler base (H.D.T. technology – Heat-pipe Direct Touch) and hold 42 aluminum plates. The heatsink plates are about 0.3 mm thick and are spaced out at a 2mm distance from one another. The heatsink measures 92 x 51 x 133 mm. The side edges of the heatsink plates are bent downwards and closed together:
The cooler base is covered with protective plastic film warning the user that it needs to be removed before cooler installation. The heatpipes in the cooler base are inserted in an aluminum plate, so that the distance between them is 3mm:
Such large gap between the heatpipes in the cooler base that contacts directly with the CPU heat-spreader will usually have a negative effect on the heat transfer and cooling efficiency of the system. At the same time, the base is very even, and its finish quality is quite fine for the coolers of this sort:
Xigmatek HDT-SD964 heatsink comes with a seven-blade 92 x 92 x 25 mm fan:
Its rotation speed is PWM controlled in the interval between ~1200 RPM and ~2800 RPM with the airflow between ~39 and 54.6 CFM. The level of generated noise in this case is from ~22 to ~34 dBA. The fan is attached on four silicone spindles fitting into the heatsink grooves:
These compensators are designed to lower vibrations and shift the sound pressure to the more acoustically comfortable range. In other words, fans run quieter with these spindles. But that’s not all yet. If you remember, the accessories bundle included not 4 but 8 silicone spindles, i.e. considering that the heatsink is symmetrical and the grooves are on both sides of it, you can improve the cooler efficiency even more by installing a second fan for air exhaust.
The assembled Xigmatek HDT-SD964 cooler with a fan attached looks as follows:
The cooler with a fan weighs 466g.
The new cooler is designed to fit all contemporary platforms except LGA 1366. To ensure compatibility with the latter, Xigmatek will soon release a new retention kit that will be included with all new coolers. Now the cooler can be installed with the common retentions of two kinds:
The detailed cooler installation instructions are provided in the corresponding manual (PDF file, 5.98 MB), although even without it, everything is simple and intuitively clear. The distance from the mainboard surface to the lower heatsink plate is 40 mm, so Xigmatek HDT-SD964 cooler will not interfere with any electronic components in the area around the processor socket.
This is what the installed cooler will look like inside a system case:
As you know, coolers with heat-pipe direct touch technology should be installed onto quad-core processors in such a way that the heatpipes go along the processor socket lock lever (photo on the right), and not across it. In this case the cooling efficiency will be significantly improved. Xigmatek HDT-SD964 is no exception from this standpoint, because when we installed it as the photo on the right shows, the temperature of the hottest processor core under maximum workload was 2~3°C lower.
In conclusion to this part of our article I would like to add that the MSRP for Xigmatek HDT-SD964 cooler is set at $34.
Now let’s check out our second testing participant, a cooler from Xigmatek, too, but from a higher-end category.
Xigmatek HDT-S1284 cooler is also a new product, although its name and design are similar to the previously reviewed Xigmatek Achilles S1284. This time, however, the box is designed in light-green color, although it still has the same information on it, such as key features of the cooler, detailed technical specifications, the list of compatible processor sockets and other less important info:
There is the same polyurethane foam casing inside the cardboard box that holds the cooler and the fan and a plastic bag with accessories stored in its upper part:
Among the bundled accessories are the following components:
Xigmatek HDT-S1284 cooler is made in Taiwan.
Now let’s take a look at the heatsink:
It looks very similar to the heatsink of Xigmatek Achilles S1284 cooler, that I have already mentioned above, however, there are a few very significant differences between them. First of all I mean the larger number of aluminum heatsink plates in the array. There are 54 of them instead of 46. Since the heatsink dimensions remained the same – 120 x 60 x 159 mm – they had to make the plates thinner (they used to be ~0.25 mm thin, and now they are only ~0.20 mm) and to reduce the gap between the plates from 2 to 1.5 mm. moreover, the plate side edges are no longer bent and closed together, like we saw by Achilles. So, the heatsink have become denser and more saturated, and its effective surface area has definitely increased:
Copper heatpipes (without any nickel-plating this time) are 8mm in diameter and are shifted from one another so that the heat could get more evenly distributed over the heatsink body.
The base has also undergone some modifications. Now the distance between the heatpipes is minimal and equals 1~1.1 mm (Achilles used to have 2 mm):
I believe it is a very serious improvement, because the smaller are the gaps between the heatpipes in the base of a cooler, the more efficiently the heat will be transferred from the CPU heat-spreader to the heatpipes. The base is impeccably even and finished for an HDT solution. The thermal compound imprint on the glass surface and most importantly on the CPU heat-spreader was practically ideally even:
As you can see, only two central heatpipes have full contact with the CPU, while two side heatpipes cover the processor heat-spreader only by half. Of course, if we compare the base of the new Xigmatek HDT-S1284 against that of the old one, the contact will look undoubtedly better on the new one.
The cooler is equipped with 120 x 120 x 25 mm fan with 7 blades:
The fan rotation speed is automatically adjusted from ~800 RPM to ~1500 RPM using PWM method. The maximum airflow it creates equals 56.3 CFM and the level of generated noise shouldn’t exceed 27.2 dBA. They used rifle bearing that should last for 40,000 hours (over 4.5 years of non-stop operation). The fan is attached to the heatsink with the same silicone spindles inserted into heatsink plates.
The cooler is installed exactly the same way as the above described Xigmatek HDT-SD964. If at some point you have any questions about installation procedure, you can always download the manual from the official web-site (PDF file, 3.92 MB).
Inside a system case the new Xigmatek HDT-S1284 looks as follows:
The distance from the mainboard surface to the lower heatsink plate is sufficient for the cooler not to interfere with anything around the processor socket. Xigmatek HDT-S1284 MSRP is set at $45.
You may remember that coolers employing heatpipe direct touch technology not only depend on the way they are installed on the CPU, but have one very serious drawback: weak pressure against the processor heat-spreader. Xigmatek seem to be aware of this issue that is why they are offering alternative retention kits for their cooling solutions (although they only come as an additional product at this time and are not yet included into the accessories bundle). These kits are called Crossbow and come to the user in a small plastic box with a cardboard insert:
If you open it up, you will get step-by-step installation instructions on using Crossbow retention with the coolers:
The actual retention consists of two steel brackets with spring-screws, a backplate for LGA 775 or Socket 939/AM2(+), a wrench and two screws:
Two retention brackets are fastened to the cooler base with the bundled screws, instead of the spinles with plastic clips:
And then the cooler is fastened to the backplate through the mainboard PCB:
As a result, the cooler is not only pressed firmer against the CPU heat-spreader, but the mainboard is also protected against bending. Crossbow ACK-I7751 retention is compatible with HDT-S963, HDT-S1283, RED SCORPION-S1283; Crossbow ACK-I7753 is designed for HDT-D1264, HDT-D1284, Achilles S1284 and HDT-SD964, and Crossbow ACK-I7754 – will suite for any Xigmatek cooler to be installed onto and AMD platform. Each kit is priced at only $7.
And finally, the third new solution based on Heat-pipe Direct Touch technology – a cooler from OCZ Technology with an ambitious name – Gladiator Max. A small box made of thin cardboard is designed in blue and light-blue colors with a cut out window in the front panel that reveals part of the cooler:
Overall, the package informational content is quite typical: distinguishing features, technical specifications, list of compatible processors and sockets and several cooler photos. It is interesting that there is the same polyurethane foam casing inside like the ones we have just seen by Xigmatek coolers. Moreover, I had a very strange “déjà vu” feeling all the way. There is a plastic bag with accessories on top including two retention brackets for LGA 775 mainboards, swing-clip for AMD platforms, silicone spindles, screws and Stars thermal compound (that is called Generic Thermal Compound for some reason):
There is also an installation manual. Neither the box, nor the manual mention where this cooler was made. We couldn’t guess its origin even from the barcode. However, this is not a problem.
Just like the previous two cooling solutions, OCZ Gladiator Max comes with its heatsink and fan packed separately. This is what the heatsink looks like:
Its general structure is similar to that of Xigmatek HDT-S1284, because it is built on four copper heatpipes 8mm in diameter that form part of the cooler base. They hold an array of 42 aluminum plates, each about 0.25~0.30 mm thick. The plates are spaced out at 2 mm distance from one another:
Each aluminum plate bears an embossed OCZ logo. Heatpipes are also shifted away from one another.
The distinguishing feature of OCZ Gladiator Max cooler is its base that has six aluminum rods piercing four bottom heatsink plates:
I have to admit that it is a pretty strange solution. If these rods should help transfer the heat from the base to the heatsink plates, then why haven’t they been made of copper, not aluminum? And if they are intended to improve sturdiness, then I really don’t understand why in this particular zone.
Anyway, let it be on the engineers’ conscience. As for us, let’s check out the cooler base:
The heatspipes in the base have been finished as good as those by Xigmatek coolers. However, the aluminum insert between the heatpipes is a little wider here and equals 1.5~1.7 mm. The base is impeccably even:
OCZ Gladiator Max comes with a 120 x 120 x 25 mm fan with 7 blades:
The fan rotation speed is controlled using pulse-width modulation approach in the interval between ~800 RPM and ~1500 RPM at 19.6~26.4 dBA noise. The specifications suggest that this fan also uses a rifle bearing with the same MTBF as the one from Xigmatek: 40,000 hours. The retention mechanism is also the same: vibration-absorbing silicone spindles:
The assembled OCZ Gladiator Max looks as follows:
Here I have to add that it weighs 780g.
It uses exact same retention kits as Xigmatek coolers:
As a result, Xigmatel Crossbow retention kit will work perfectly fine on OCZ Gladiator Max:
In conclusion I would like to show you two photos of the OCZ Gladiator Max cooler installed into the system case:
Once again, when we installed the cooler with the heatpipes along the processor socket locking lever (photo on the right) the cooling efficiency improved by 2°C. The distance from the mainboard surface to the lower heatsink plate of the cooler measures 41mm.
OCZ Gladiator Max is priced at $45.
The technical specifications and recommended retail price of our today’s testing participants are summed up in the table below:
We tested the new coolers and their competitor 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.
Our testbed was identical for all coolers and featured the following configuration:
* - some photographs have been taken with Radeon HD 4870 graphics card installed; however, the tests have been performed with GeForce GTX 260.
All tests were performed under Windows Vista Ultimate Edition x86 SP1. SpeedFan 4.36 beta 15 was used to monitor the temperature of the CPU and mainboard chipset, reading it directly from the CPU core sensor and to monitor the rotation speed of the cooler fans:
The mainboard’s automatic fan speed management feature as well as CPU power-saving technologies were 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 in two modes. First we used Linpack 32-bit with convenient LinX shell version 0.4.9 to heat up the CPU to its maximum. We manually set the RAM capacity at 1850MB and recorded 15 runs.
Since we ran the test twice with 20-minute idle period between the runs for the system to cool down and temperatures to stabilize, the relatively short actual testing period was quite enough for the maximum processor temperature to become stable.
Since Linpack loads the CPU in a unique way (very high load), we also performed additional tests with more reasonable load in a popular Far Cry 2 game. We ran the built-in benchmark in “Ranch Long” mode three times with “Very High” settings (DirectX 10) in 1280 x 1024 resolution:
According to the benchmark results, the average frame rate was at 93~95 fps. I replaced Unreal Tournament 3 with Far Cry 2 in my tests because Far Cry 2 loads the CPU more, which gives us higher temperature of processor cores.
I performed at least two cycles of tests in both test modes 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. We took the maximum temperature of the hottest processor core after two test cycles for the results charts (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 23.5~24.0°C. It is used as a starting 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 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 3cm, 1m and 3m distance from the noise source. During the acoustics tests all three 120-mm case fans were slowed down to ~700 RPM. In this mode the background noise from the system case measured at 1m distance didn’t exceed ~32.8 dBA, and the loudest fan was the 140-mm fan of the system power supply. 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 34~34.5 dBA.
For the reference purposes we also included the results for Thermalright Ultra-120 eXtreme ($60) tower cooler equipped with a PWM fan from Xigmatek HDT-S1284 rotating at 860~1460 RPM:
Moreover, all coolers except the small Xigmatek HDT-SD964 were also tested with a 140 x25 mm Scythe Kaze Maru fan ($20) working at its maximum rotation speed of ~1830 RPM:
We didn’t install the second additional fan onto Xigmatek HDT-SD964, because I didn’t have it available at the time of tests, and the 80-mm fans that I did have, simply couldn’t be installed onto its heatsink the regular way.
During Linpack tests inside a closed system case using the “weakest” cooling system of the today’s testing participants we managed to overclock our 45 nm quad-core processor to 3.7 GHz (+23.3%). The nominal processor Vcore was increased to ~1.4875 V in the mainboard BIOS (+29.3%). During the tests in Far Cry 2 game the CPU remained stable up to 3.95 GHz (+31.7%) at 1.525V (+32.6%) Vcore:
Far Cry 2
Before we start discussing the obtained results, let me make a few things clear. So, all tested cooling solutions employing Heat-pipe Direct Touch technology were installed in their most efficient position, i.e. with the heatpipes directed along the processor socket locking lever (parallel to the power supply unit). Then we found out during the test session that we could win 2°C under maximum workload in Linpack by simply replacing the default plastic clip retention of Xigmatek HDT-S1284 and OCZ Gladiator Max coolers with the Crossbow kit. However, Xigmatek HDT-SD964 didn’t show any improvement with the replaced retention. Nevertheless, the results on the diagrams below are given for all three coolers tested with Crossbow retention instead of their default one.
The efficiency tests are summed up on the diagram below:
Xigmatek HDT-SD964 being the last in this race was pretty predictable. Small heat dissipating surface and wide aluminum inserts between the heatpipes in the base of the cooler do not let this solution catch up with the leading group. However, I have to say that HDT-SD964 only lost 2°C to the next contestant – OCZ Gladiator Max, though the latter is way quieter. Frankly speaking, Gladiator didn’t impress us with its efficiency. Its heat dissipating surface may be too small for this cooler to reveal its potential to the full extent. With a powerful Scythe Kaze Maru fan OCZ solution gets 3°C more efficient inside a system case, and 4°C more efficient in an open testbed.
And the third cooler with heat-pipe direct touch – Xigmatek HDT-S1284 – turned out the best of the three solutions tested today. It is not dramatically ahead of the other two testing participants, but it is still ahead. High heatpipes density in the base and large effective heatsink surface help Xigmatek solutions win this race. With a 140-mm Scythe fan this solution runs as efficient in an open testbed as the Thermalright Ultra-120 eXtreme (with the same fan). Not bad, considering that Xigmatek HDT-S1284 is about $15 cheaper and is already bundled with a fan, while Ultra-120 eXtreme comes without one, so you will have to purchase it separately.
Now let’s check out the acoustic performance of our today’s testing participants:
As you see, the coolers’ noise levels do not go beyond the subjective comfort zone only with the fans at their minimum speed. During high CPU utilization when the fans speed up to their maximum, the coolers become very noticeable against the background of a quiet system case. Moreover, Xigmatek HDT-SD964 is way too loud.
In my humble opinion, the most significant part of our today’s article is the fact that manufacturers started paying attention to the comments about weak pressure of heat-pipe direct touch coolers against the CPU heat-spreader. Now you can find a reliable retention kit for your cooler that will ensure highly secure contact. Of course, it would be best if these kits were included with the bundled accessories of Xigmatek HDT-S1284 and OCZ Gladiator Max coolers. I don’t think it is going to affect their price too much; however, it will definitely improve their cooling efficiency by a few degrees. And let’s not forget about the retention for LGA 1366 processors that would be nice to have and that is not yet bundled with Xigmatek and OCZ coolers.
As for the specific cooling solutions discussed today, I have to admit that they are not too exciting by today’s standards. Xigmatek HDT-SD964 can catch your eye due to compact size and low price of only $34, but at the same time it may discourage some potential users by its unimpressive cooling efficiency and noisy operation. You may be able to improve its performance by adding a second exhaust fan to it, but as a rule, the improvement is no more than 3-4°C. OCZ Gladiator Max cooler proved a little more efficient, however with a competitor like Xigmatek HDT-S1284 priced identically, the Gladiator may have hard times finding new home. Xigmatek HDT-S1284 at the tested fan speeds is just a little behind Thermalright Ultra-120 eXtreme, but with a high-performance fan these two solutions run equally efficient.