Cooling has become one of the most important aspects of a computer's design. With newer CPU's emitting more and more heat, so too the cooling systems for these have had to keep up. Many companies have relied on one of a couple of techniques, a fast high volume fan which in turn is very loud; a large heatsink design to transfer more heat away; or a copper based cooler. Many companies use a combination of these techniques, mainly using copper and a loud fan to cool down the processor.
In the past two years a company called came onto the cooling scene with a heatsink called the SK6. Many sites wondered whether this new company could compete with the likes the of Alpha PAL series, the Swiftech MC series, and the Global Win heatsinks. This heatsink proved itself as a match for the then champion Swiftech MC-370 series heatsink in many reviews, and gathered much deserved praise. The design was fairly simple and incorporated two of the techniques mentioned previously, by being all copper and using the infamous 60mm Delta fan that many people loved to hate. But since that heatsink was released early last year other companies have released newer and better products. Has Thermalright been sitting around since they released this product? Well let us look at what they have released since then.
We will be looking at two heatsinks from Thermalright today, and I would like to thank for providing these products for us to test. First we will look at the SK6+ and then the SK7. So let us see how the SK6+ looks first.
The SK6+ is, as the name suggests, an update of the very successful SK6 heatsink. How is it different from it's older brother? And how does the packaging look, is it flashy or just a standard box? We will let the pictures themselves answer these questions. But for a more detailed specification of the heatsink please look at .
The box that the SK6+ comes in is very plain but it still manages to protect everything very well, with nothing being really loose. The second picture shows the clips that attach to the base of the SK6+ (more on that in installation), and has clips for both 25mm and 38mm thick 60mm fans. Also included with the heatsink is a clip which uses all of the 'lugs' that the socket provides (three on each side), theoretically giving the heatsink a better grip on the processor making it less likely to fall off if one lug breaks.
There are 37 soldered fins on each side of the heatsink for a grand total of 74 copper fins connected to the copper base by means of solder. This, in comparison to many newer heatsinks which are skived, could result in a lower cooling performance if the solder joints are not connected very well. The all copper heatsink helps bring the total weight of the heatsink up to 382g without a fan, which is well above AMD's recommended maximum weight for a Socket A based heatsink.
The size of the heatsink is also pretty average at 74 (L) x 65.5 (W) x 33 (H) mm. Also included is a 1ml plunger of thermal compound which I didn't use in testing but which works fine on my newer PIV CPU. The base of this heatsink is very reflective as the last picture shows, and it comes relatively unblemished due to the plastic bag the heatsink was in, inside the box.
No fan was included with this particular heatsink but you can add any 60mm fan that you want with it (there is even a space for a fan in the box), either before you purchase it or from one you have lying around. Overall the SK6+'s design is nothing extremely new, but the three pronged clip and very shiny base help raise the design above just normal. What about the SK7, is it a unique design or is it similar to the SK6+?
How is the SK7 different from its sibling, the SK6+? Is it a totally different design, perhaps a skived copper heatsink? Once again we will let the pictures speak for themselves, but to look at Thermalright's specifications of this heatsink, .
This heatsink is based on the same basic idea as the SK6+ but there are some very interesting differences between the two. The most obvious of the differences is that the SK7 is bigger, with dimensions of 85 (L) x 65.5 (W) x 42 (H) mm without fan, which is 11mm longer, and 9mm taller than the SK6+ and as a result is 103g heavier than the SK6+ (485g). It uses the same technique for attaching the fins as the SK6+, as the fifth picture shows. However its increased length allows it to be able to have one very interesting and useful feature, that of being able to use any 60, 70, or 80mm fan that you want, on the same heatsink. I believe that this is a good idea as it allows you to attach your old 60mm fan until you get your 80mm or even 70mm TMD fan.
The clip used is the same one as on the SK6+ but it is unable to be removed without damaging the actual heatsink. The 34 fins, which cover the entire length of the heatsink unlike the SK6+ help draw the heat away from the 10mm thick base. The base of this heatsink was very similar to the SK6+ which means it was very nice, and it came protected with a thin piece of clear tape with a red edge to it (remember to take this off before putting this on your CPU). The box that the SK7 came in was larger than that of the SK6+, but was packed even more with clips (for 60, 70, 80*25mm fans and 80*38mm fans), the same amount of thermal compound as well as the heatsink in a plastic bag.
This heatsink was also sent to me without a fan, but also included in the shipping box that I received that held the heatsinks was one of Thermaltake's SmartFan II's, a 80mm heatsink that Hubert looked at in his Volcano 9 review.
The heatsink installed onto the motherboard very easily, obviously Thermalright has put some thought into the design of the design of the clip to make installation problem, and slippage free. Installing the fan however was another matter. The clips that are included to hold the fan on are very tight, which is a good idea, however they are too tight, as I was not able to install a fan grill on top of the 7000RPM delta fan. This is not a good design plan as if you cannot put a grill on top of the heatsink, you may end up cutting off a finger or having some of your cables cut, such as power and IDE cables. Un-installation was not as easy as the installation, as it required two tools to take it off.
The installation of the heatsink onto the motherboard was a fairly problem free action. Using a flat-head screwdriver I was able to get the heatsink installed, with only the one tool. Un-installation was more of a problem as I needed both the screwdriver and another tool to angle the clips outward. Installing the fan was similar to that of the SK6+ though the selection of clips will confuse some.
What was the system used to test these heatsinks, and what did we use to compare the SK6+ and SK7 too?
||AMD Athlon XP 1800+ (TBred Core) - 1.39-1.46v (37-43Watts)
MSI 745Ultra - SiS745
||512MB PC2700 RAM
| Drives (In Order: Top-Bottom):
MSI Dragonwriter 40/12/48, 4.3GB Quantum SE, 40GB Maxtor 7200RPM (D740X)
||Matrox Parhelia - 200/250MHz
||Windows 2000 Professional Service Pack 3
||Parhelia Driver - 1.03.00.043
||Firewire card, D-Link 538TX NIC
||Thermalright SK6+, Thermalright SK7, Zalman CNPS3100-Plus, AMD Stock Heatsink (XP1800+)
||Thermaltake Smart Fan II (Medium/High), 60*25mm Delta Fan (Only on SK6+), Stock AMD 60*15mm Fan (Only on AMD Heatsink)
||InWin Q500 - 1-Rear Generic 80mm Fan; 1-Front 80mm Fan
||Enermax EG365P-VE 350Watts
|| Zalman fan bracket
||Motherboard Monitor 5.02, MBMLog, SiSoftware Sandra 2003 Burn In tests.
The heatsinks were tested using an electrical thermal compound (seen below) called Thermalcote™. The Smart Fan II was only tested using the manual fan control set at a halfway point and then at the maximum fan speed. Both fans used the 3 pin splitter to 3 and 4 pin splitter so that we could monitor the fan's RPM while not drawing any power from the motherboard.
All the tests were run for about 15 minutes of activity, and 15 minutes of inactivity. This was recorded using MBM and MBMLog to create a graph of the CPU/Case temperature during the testing. All of these tests were done at least twice and more if the results were not within a 1C error range. CPU temperature was read using a sensor underneath the socket by Motherboard Monitor. Each heatsink was tested using the Smart Fan II and the Zalman fan bracket. The SK6+ was also tested using the Delta 'black label' fan on top of the actual heatsink and the stock AMD heatsink used just the stock fan.
Sound tests were done from the closed case with the sound meter being about 18" (45cm) from the case. The sound meter was a , which was running at the lower limit of its range in these tests. So how did the Thermalright heatsinks fair, did they perform much better than the stock AMD heatsink or the Zalman heatsink? Lets look at the results.
Next Page - Testing and Conclusions