by Sergey Lepilov
09/07/2005 | 04:17 PM
The heat dissipation of contemporary processors increases with the release of every new model. The cooling solutions for the new CPUs also tend to become more and more sophisticated, because the cooling requirements grow almost linearly, and the use of heatpipe technology has nearly turned into a de-facto standard already. As of today, almost every well-known CPU cooler manufacturer offers solutions based on heatpipe technology. Among them are the systems from Gigabyte, ASUS, CoolerMaster, Scythe and many others have already been reviews on our site. And even the aluminum guru, GlacialTech Company, released their Igloo 5600PWM cooler (although only for LGA775 processors yet), which is also built around heatpipe technology.
At the same time there is a lot of discussion going on about the today’s most efficient cooling solutions, including those using heatpipes. We have got a lot of requests from you, our readers, to gather all the cooling monsters together for a single test session, so that we could finally single out the winner.
I would like to point out right away that at this point we are not planning a large-scale testing session like that, however, we are going to introduce to you three very interesting representatives of the cooling systems of this type and compare their characteristics and performance against the results of the recently reviews Zalman CNPS9500 LED (for more details check out our review called First Look at Zalman CNPS9500 LED: the Power of Air, the Efficiency of Water ). We will also include the results of a reference boxed cooler for AMD Athlon 64 3000+ (Socket 939) processor.
So, today we will be talking about two members of the Titan cooler family: Vanessa S-type and Vanessa L-type, and one really big cooling solution from Thermaltake : the Big Typhoon. In our article we will not only describe the coolers, discuss their design peculiarities and technical characteristics, but also test their CPU cooling efficiency and estimate the level of noise generated during the system operation.
Well, please meet our today’s testing participants now.
The cooler we are going to begin with is shipped in a very beautiful box made of thick carton. The box is very large, I would even say unusually large for a package of a cooling solution. The box dimensions are 26.5cm x 20cm x 10cm (LxWxH). On the front cover there is a pretty girl with butterfly wings, which look like they have been made of some silverfish metal.
There is also a small photo of the Titan Vanessa S-type cooler and a few icons highlighting the major cooler advantages from the manufacturer’s point of view: high CPU cooling efficiency, low noise level, simple installation procedure, Titan Nano Blue thermal paste included into the bundle and fan rotation speed.
In my humble opinion, the reverse side of the package seems to be much more interesting and informative:
There you can find detailed specifications and cooler characteristics, a large photograph of the actual cooler with component listings and their detailed descriptions, a list of supported CPU form-factors and a photo of the fan rotation speed manager. On the sides of the box there are more cooler pictures taken from four different sides and a complete list of all the accessories bundled with this cooler that you will find inside the box. Of course, there is the usual “must have” info about the product meeting ecological standards, the bar-code and the URL of the manufacturer’s official web-site.
Inside the carton box there is another transparent plastic box with a cover. All the Titan Vanessa S-type cooler components are placed in their designated sections:
Under the fan there is a small box with the retention mechanisms for all supported CPU socket types and a tube with thermal paste:
The accessories bundled with the Titan Vanessa S-type cooler include the following:
If you have been reading our cooler reviews you should already be familiar with the design of Titan Vanessa S-type cooler. It is built of aluminum plates stacked onto three heatpipes:
These heatpipes are made of aluminum, not copper. There are 45 aluminum plates, each ~0.5mm thick. They have larger contact surface with the heatpipes than just their cross-section: when the holes in these plates were squeezed, there formed a kind of small aluminum skirt around the heatpipes, which is a little bit bigger than 0.5mm. The heatpipes are ~5mm in diameter. The entire heatsink construction without a fan is very light-weight, despite its impressive size. The ends of the heatpipes are covered with blue rubber caps, which serve not only some protective purposes preventing the heatpipes from physical damage, but also add to the aesthetic looks of the cooler.
Titan classified Vanessa cooler family as “butterfly design”. And this is certainly a very logical name. Take a look yourselves:
The aluminum plates are really shaped like a butterfly. Moreover, the very top plate is covered with a decorative sticker of a butterfly-girl from the fantasy land (I assume her name is Vanessa :) ).
The fan is designed in a blue-silver gamma. It is 92mm x 92mm x 25mm in size, is covered with protective grid and features seven silver blades:
The fan can be installed onto the heatsink only on one side of it. It is fastened with two metal brackets, which should catch to the corresponding slits in the heatsink plates:
To tell the truth, this retention solution seemed to be very unstable and unreliable at first glance, however I have to admit that even when the fan was working at its maximum rotation speed, I didn’t notice any excessive vibrations.
The cooler foot is covered with protective plastic film sticker reminding the user that this sticker should be removed before installing the heatsink onto the CPU:
The surface quality is truly ideal. I don’t think I have to say how perfectly flat it is, but as for the polishing quality, you can see it yourselves from the picture above. By the way, Titan claims in the cooler specification that the heatsink foot is made of solid copper and features platinum shine finish (it is most likely to be the usual nickel-plated finish).
The part of the heatpipes contacting the cooler foot is slightly squeezed. There is special thermal interface applied between them and the heatsink foot, to ensure better contact:
The bottom plate is pressed against the top bracket with the regular screws. In the center of the bracket there is a square lug. It ensures better fixation of the cooler retention brackets that press the cooler against the CPU and prevent it from moving sideways:
The cooler can be installed onto the CPU very easily. It is a totally intuitive process. That is why we will not go into details regarding the cooler installation. For those who may still have questions here is a link to the official Titan page where you will find installation tips for all supported processor sockets. The most important thing that is worth stressing is that you install the heatsink onto the CPU first, and only then fasten the fan to it.
Once assembled and installed, the Titan Vanessa S-type cooler looks as follows:
As you can see, there is one unpleasant issue about it. When you install the cooler onto an LGA775 or Socket 478 platform, you can turn the heatsink and fan facing in any of the four directions. In case of a Socket 754/939/940, if the fastening holes on the mainboard PCB are located vertically (which is typical of most K8 mainboards), you can only choose from two possible installation positions. Namely, you can either direct the air flow from the graphics card towards the system PSU, or in the opposite direction (is there is enough room between the edge of the mainboard PCB and the system power supply unit). None of the options is actually acceptable for efficient cooling of the system interior. The warm graphics card being so close will hardly allow fresh cool air to get to the CPU fan. The close location to the system PSU also won’t do the CPU cooling any good. The heatpipes and the retention design will not allow you to turn the cooler so that the air flow could go directly to the case fan to be sent outside. However, if you are a lucky owner of a K8 mainboard with horizontal location of the cooler retention holes, you will not have to bother with this problem.
In conclusion to our installation discussion I would like to add that the fan rotation speed controlling device can fit not only into the 3.5” bay, but also onto the case rear panel instead one of the brackets (in this case you will have to remove the face panel from the device in advance). It is totally up to you to decide which of the two ways better serves your needs.
Vanessa L-type required a larger box: 33cm x 19cm x 13.5cm (LxWxH), however its design and colors are very similar to those of Vanessa S-type package:
On the back of the box we see pretty much the same technical details, with that only difference that the specifications and the picture belong to L-type cooler model:
Inside the box there is the same transparent plastic package with very similar contents:
Besides a long screwdriver, there are the following components included into the package:
Titan Vanessa L-type is a heatsink with only one central heatpipe 25mm in diameter. The heatpipe is made of copper and has looping internal structure (according to the manufacturer). There are 34 aluminum plates forming the heatsink fins around the central heatpipe, each is ~1mm thick. The 120mm fan is fastened with screws to one of the sides of the heatsink. The entire construction looks very nice:
The top plate of the heatsink is decorated with the picture of a batter-fly girl Vanessa:
The fan of Titan Vanessa L-type is designed in exactly the same way as that of the S-type cooler we have just discussed. It is big enough to cover all the heatsink plates completely:
Here are more cooler pictures at different angles, that will give you better idea of its appearance:
The round sole of the cooler is protected against physical damage that may occur during transportation with a blue plastic sticker:
The cooler sole is very well polished and even though there no mirror shine, you can hardly feel the machining marks by touch.
The cooler installation onto any CPU socket begins with fastening the metal frame with the round hole for the copper cooler sole:
Then you take the screwdriver included into the cooler bundle and fasten the cooler to this frame (don’t forget to apply some thermal paste first). Due to the symmetric location of the fastening holes this cooling device can be installed in any position, so that the air flow is directed in the most optimal way for better cooling efficiency. This is a definite advantage compared with Vanessa S-type.
The cooler looks very beautiful inside the system case:
Although we still had to move the memory modules into another channel, because the DIMM installed into the first memory slot pushed too hard against the cooler fan. Other than that we didn’t have any more installation problems with Titan Vanessa L-type and our ASUS A8N-ALI mainboard. The detailed installation tips for all socket types are available here .
We have already tested this great cooling solution and discussed most details about it in our article called Four CPU Coolers from Thermaltake Tested . However, in case you missed that article we will repeat a few most essential things here.
The transparent plastic package allows you to take a closer look at the cooler even before you buy it:
There was no room on the face cover of the package for any technical info and specs, so Thermaltake placed there just three major specification highlights a few small pictures of the cooler. The reverse side of the package also doesn’t reveal too many technical details:
There are six icons-pictures of the cooler components and the bar-core. The detailed technical specs are listed on one of the sides of the pyramid-shaped package.
All the bundled accessories are packed in a black carton box at the bottom of the package. Namely you will find the following items inside this box:
As you see, there is no fan rotation speed control device included. Although this cooler doesn’t really need it.
Thermaltake Big Typhoon is designed with 6 copper heatpipes (two sets, three heatpipes in each). The diameter of these heatpipes is about 6mm. They are curved to a trapezium shaped construction ending at the top of the cooler:
The heatpipes go through 140 aluminum plates, each 0.3mm thin. The aluminum plates are also split into two sets, 70 in each set. The top of the whole construction is covered with metal casing with Thermatake’s brand logo on the sides, and the 120mm fan is fastened to the casing:
The orange fan blades are covered with protective grid. The fan rotation speed is 1,300rpm, according to the specifications. Besides, Thermaltake offers the users a few optional fan designs. The standard Big Typhoon fan can be replaced with a fan featuring rotation speed control (A2029 ) or a fan featuring rotation speed control and decorating LED lighting (A2018 ). This way you can improve the functionality and exterior looks of the device. The power cable of the fan is hidden inside a soft plastic sleeve. A very pleasing trifle, I should say.
The heatpipes sit in special grooves in the heatsink sole, so that the contact area gets much bigger than if they were just lying on the flat surface of the sole. So, the heat dissipation goes more efficiently than by Titan Vanessa, for instance:
The top of the heatpipes is covered with a copper plate, which also supports the cooler retention mechanisms.
The sole surface is processed very poorly. Despite the seemingly flat surface, you can clearly feel the machining tracks by touch:
And don’t even mention any polishing or mirror-shine; they are not there. This is probably the only drawback about the design of Thermaltake Big Typhoon cooler. Besides, you may also notice some scratches left by the CPU covers (we have been using this cooler in our lab for a month already), but this is certainly not a design drawback.
The cooler can be easily installed onto any CPU socket. There is a long top bracket and two different back-plates that make the entire cooler retention. Once the cooler has been installed onto our ASUS A8N-SLI, the whole thing looks as follows:
Everything seems to be pretty simple at first glance, however, it is very inconvenient to fasten the screws because of the top aluminum fins block. You have to hold the screwdriver at an angle, it slides over the head of the screw, and the L-shaped screwdriver was not included with the accessories set. Moreover, if you are installing the system into a standard ATX case you will have the power supply unit in your way. Here are the photos of the installation guide for your reference: front page and back page .
Thermaltake Big Typhoon looks like a real giant inside the case:
In conclusion I would only like to add that Thermaltake Big Typhoon is the only solution of all our today’s testing participants that generates a very efficient airflow inside the system case: it cools all the electronic components located around the CPU on the mainboard and even the system memory.
Once again, if you would like to get more details on Thermaltake Big Typhoon and other Thermaltake coolers, you can check out our recent article called Four CPU Coolers from Thermaltake Tested.
Let’s sum up all the specifications of our today’s testing participants. For a better comparison we have also added here the specs of the recently reviewed Zalman CNPS9500 LED, which will compete in our test session today. For more details on Zalman solution please check out article called First Look at Zalman CNPS9500 LED: the Power of Air, the Efficiency of Water .
Well, since we discussed the Zalman cooler in great detail already, we would only like to show you its picture inside our testbed:
In the dark the LED’s of the cooler look even more beautiful, however, against the background of two Sharkoon Luminous Blue LED fans Zalman CNPS9500 LED looks quite modest, I should say:
As we can see, I managed to turn Zalman CNPS9500 LED so that the air flow could be directed towards the case fan. In order to turn the fastening bracket in the appropriate way, you need to slightly push up the lower heatsink fins. In this case you will easily mount the cooler in any way irrespective of the fastening holes location around Socket 754/939/940.
Besides the coolers listed in the table above, we have also included the performance results for the regular boxed cooler supplied with AMD Athlon 64 (Socket 939) processors. Here are a few photos to remind you what this baby looks like:
The fan of this cooling device can rotate at speeds between ~3,000rpm and ~4,800rpm depending on the CPU temperature (it wouldn’t rotate any faster in our testbed). The generated noise is pretty loud.
Let’s take a look at the testbed configuration before we pass over to analyzing the obtained performance results. Especially, since there have been a few important changes in our testing approach that are worth mentioning.
All the tests were carried out in the following closed testbed:
All tests were performed in Windows XP Home Edition Service Pack 2 with NVIDIA nForce system driver version 6.39, DirectX 9.0c and Catalyst 5.8 video driver.
It is quite logical that you don’t really need these powerful coolers if your AMD Athlon 64 3000+ works at its nominal speed of 1800MHz, so their working efficiency would be limited by the low CPU heat dissipation. That is why we decided to eliminate this limiting factor and overclocked our AMD processor to 2713MHz with 1.55V Vcore setting:
We used S&M utility version 1.7.1 for CPU warm-up, temperature monitoring and fan rotation speed control. We warmed up the CPU for 15 minutes in the Normal mode under 100% workload:
Moreover, keeping in mind that S&M loads the CPU very heavily, which is not typical of most software applications (I would even say that this workload is truly one of a kind), we performed the CPU warm-up tests with a Super PI benchmark, which is very popular among overclockers today. The calculation of the PI value up to 32 million digits takes about 26-27 minutes on the platform like ours.
All the coolers were tested in exactly the same testing conditions in the closed system case. We ran the tests in two modes: with both 120mm case fans on, and without them. The room temperature in our lab stayed at 22oC, which is our reference point on all diagrams.
Let’s first take a look at the results obtained in the most favorable operation mode, when both case fans are working:
* - the fan rotation speeds are taken not from the specification list, but are calculated as the average rpm value according to S&M measurements.
Well, let’s start our discussion of the obtained results.
The standard boxed cooler from AMD Athlon 64 (Socket 939) appeared surprisingly powerful and managed to retain the temperature of the overclocked processor within acceptable limits. The bad thing about it was though that the temperature growth pushed up its rotation speed too, so that the fan rpm value increased from the nominal ~3000rpm to ~4800rpm, and the cooler starts generating pretty unpleasant noises.
The youngest heatpipe butter-fly, Titan Vanessa S-type didn’t perform that well at the minimum fan rotation speed. It lost about 7o C(!) even to the regular boxed cooler. Yes, Vanessa’s fan is practically noiseless at 1,380rpm rotation speed, but you have to sacrifice low temperature for the sake of noiseless operation. Once the fan speeds up to its maximum, Titan Vanessa S-type regains its power and the cooling efficiency improves significantly. However, it still managed to win only 3o C from the boxed cooler under the peak workload, which is unacceptable luxury for a 36-dollar cooling solution. Unless you are ready to pay for the exterior design when shopping for a CPU cooler…
Titan Vanessa L-type appeared much more efficient than its younger sister. The remarkable thing however, is that the gap between these two solutions reaches 5-6o C in quiet modes, while in the peak modes it closes to only 2o C, though the leadership still stays with Vanessa L-type.
Thermaltake Big Typhoon… what can we say here? This “natural phenomenon” demonstrates outstanding cooling efficiency for our overclocked to 2.7GHz AMD Athlon 64 3000+ processor with 1.55V Vcore. This cooler managed to get 12o C ahead of the best from the already discussed solutions under peak workload. Noiseless operation and maximum efficiency – these are the key advantages of this solution. The temperature didn’t rise above 48o C throughout the entire S&M test session (you can take a look at this graph for more detailed report). Those of you who like to compare the testing results obtained in different test sessions and described in different articles should remember that it is not always correct. Such parameters as room temperature, system case model, case fan size (120mm instead of 80mm) will certainly affect the results. Moreover, ASUS ASCOT 6AR2-B case we used this time has a grid embedded into the side panel that allows more fresh air inside the case. Far not all the case models we use for our cooler test sessions can boast something similar.
And what about Zalman CNPS9500 LED? Unfortunately, I have to admit that our hopes didn’t come true this time. :( On the one hand, it is very sad because I was looking forward to this particular cooler like many other overclockers, and it was destined to turn into a new example of high efficiency, quiet operation and great looks in my system case (the latter is purely subjective, of course). But on the other hand, things are not so bad at all, as there is a more efficient cooling solution than Zalman CNPS9500 LED! What else could we say about an 11o C difference under the peak workload when the testing participants were working in quiet mode? Yes, the 92mm fan of Zalman CNPS9500 LED running at ~1,420rpm is a little bit quieter than the fan of Thermaltake Big Typhoon running at ~1,370rpm. However, both these solutions are nevertheless so quiet, that you can only tell the difference by ear if you disable both case fans and install an ATI Radeon X300 graphics card with a fanless passive cooler onboard. I do not exclude the possibility that in case of hotter processors, such as Intel Pentium 4 Prescott @ 4.2GHz the performance difference between these two cooling solutions may grow smaller, but I am pretty sure that Thermaltake Big Typhoon will not lose. Here I also have to add that when we turned the Zalman CNPS9500 LED cooler so that the air flow was going towards the case fan instead of the system PSU, the CPU temperature got only 1o C lower. I assume that this difference is pretty insignificant because there is an 80-mm fan inside the PSU that sucks the air in from the area around the CPU cooler.
Now let’s run all the tests once again but this time without any of the case fans working. This way we will create worse working conditions for all the system components and CPU in particular:
Well, we do not see any dramatic changes here. Titan Vanessa S-type turned out so weak that it failed to ensure proper cooling for our CPU under the extreme workload created by S&M utility: system would go into BSOD mode. The boxed cooler looked quite fine against the background of the Titan Vanessa S-type, although generated too much noise during operation. The performance gap between Thermaltake Big Typhoon and Zalman CNPS9500 LED did grow smaller, although the Zalman solution still got defeated by the Big Typhoon. Another advantage of the winner is the air flow course: it blows the air onto the mainboard cooling down the electronic components around the CPU, while the Zalman solution simply ousts the air outside the case.
Now a few subjective words about the noise level. When the coolers work in the “quiet” mode, i.e. at their minimum fan rotation speed, they are all very quiet. Thermaltake Big Typhoon also belongs here, as you do not hear its fan at all against the background of the case fans and the graphics card cooler. As for the noise level when we have cooler fans running at the maximum rotation speed, we cannot call them loud too. We have to admit that at higher rotation speed you can actually hear Titan Vanessa S- and L-type and Zalman CNPS9500 LED against the background of the system case fans, however, even after two hours of constant operation you do not really get tired of the generated noise. If we line up the coolers according to the level of noise they produce, we will get the following picture: the loudest will be Titan Vanessa S-type at ~2680rpm followed by Titan Vanessa L-type (~2080rpm) and Zalman CNPS9500 LED (~2500rpm).
As we have already mentioned, Titan coolers and Zalman CNPS9500 LED go with a special fan rotation speed controller device that allows you to find the most optimal balance between noise and cooling efficiency for your particular case.
For our tests in an open testbed (outside the system case) we took Gigabyte GA-K8N Ultra-9 mainboard (BIOS F6). We used the same AMD Athlon 64 3000+ but working at higher Vcore setting of 1.7V (1.55V is the maximum ASUS A8N-SLI mainboard allows).
The lab room was 16 sq.m big. The room temperature during the entire test session remained at 21oC and was monitored with a thermometer. The mainboard was set horizontally.
We increased the CPU warm-up cycle from 15 to 70 minutes with the same S&M utility. According to the thermometer, the room temperature didn’t get any higher throughout the entire test session.
Here are the results:
Unfortunately, SpeedFan (version 4.23 and the latest version 4.25) would freeze on this mainboard at the sensors initialization stage that is why we used Gigabyte Easy Tune 5 to monitor the temperature.
Zalman CNPS9500 LED (click to enlarge)
Thermaltake Big Typhoon (click to enlarge)
The result obtained for Zalman CNPS9500 LED at 1900rpm is not accidental. The thing is that according to our subjective noise level estimate, Zalman CNPS9500 working at 1900rpm and Thermaltake Big Typhoon working at 1360rpm generate the same level of noise. At 2500rpm Zalman cooler is much louder.
Our open testbed tests allowed us to get to know Thermaltake Big Typhoon a bit better. Firstly, even under the maximum processor workload the heatpipes remain cold between the cooler sole and the first aluminum fins. However, the segments of the heatpipes right before the aluminum fins and right above them on top of the heatsink are hot, not just warm. The heatsink fins feel more like warm. Secondly, a have to correct our statement made above about the Thermaltake Big Typhoon also cooling the electronic components on the mainboard around the CPU. It does cool these components, but only in the idle mode . Under the workload such as during Super PI or S&M, the air flow coming down from the heatsink fins is warm and cannot serve any cooling purposes in the open testbed (not to mention the closed system case).
So, it looks like Zalman CNPS9500 LED may have one very last chance to regain its positions – Prescott. Will the efficiency of Zalman CNPS9500 LED in a Prescott based system make up for high price and high noise level?
Our today’s testing participants are surely very interesting solutions. Each of them has its own highlights, of course. When Titan worked on their Vanessa S-type solution they seem to have simply forgotten about the major function of a CPU cooler and focused primarily on the external looks. They should have added LEDs into the fan to make the picture complete, and then, maybe, some commencing modders-overclockers would like to get one of those coolers for their system case with a transparent side panel. However, I personally felt really uncomfortable that a 36-dollar cooling device was defeated by a regular boxed cooler.
Titan Vanessa L-type is a little better solution, however, if Titan’s competitors roll out a few hits like Thermaltake Big Typhoon, which are even better value from the price prospective, then Titan Vanessa L-type will simply have no chances. You can always purchase the same fan rotation speed manager device with additional LED lightning for something about $10.
As for Zalman CNPS9500 LED, there is nothing to add here so far. It is more efficient than the boxed cooler and is equipped with the fan LEDs from the very beginning, but it cannot beat the leader still.
I cannot tell you right now if Thermaltake Big Typhoon is going to be better than Scythe Ninja Heatpipe, because we haven’t completed our tests yet, but so far it looks like the overclocker’s choice for the next half a year has been made: if you go with air cooling - go for Thermaltake Big Typhoon.