Six Premium Heatsink Shootout

I have been involved with overclocking my systems since the Pentium MMX days. One of the first things I learned even way back then was that, if you wanted to push your system faster than Intel or AMD stock speed specifications, cooling is most important in getting stable overclocks and stock heat sinks aren’t remotely adequate for cooling when pushing the cpu clock speeds. So I started out with looking into aftermarket coolers for those systems and haven’t looked back since. Premium aftermarket coolers have been around a while and have gone through several changes since those days, until most high end air cooling solutions you find nowadays are heat pipe based solutions that generally are mounted vertically in relation to the motherboard and exhaust towards the rear case exhaust fans. I will be looking at 6 different heat pipe cooling solutions with this review, with retail prices ranging from the mid-$30 range to $100. All have been looked at by other sites at one time or another, but generally are tested on a dedicated test stand setup. I will be a bit different as I am going to test them in a case on a working system and see if any surprises will be found.

The Heat Sinks

The 6 heat sinks I will be testing are as follows:

• Thermalright TRUE Revision C 1366RT
• Thermalright TRUE Copper
• Thermalright Venomous X
• Cogage True Spirit
• Prolimatech Megahalems
• Noctua NH-D14

Packaging

Of the above six heat sinks, three of them are directly marketed under the Thermalright name and the Cogage True Spirit is manufactured and sold by Cogage, a subsidiary company of Thermalright. Thermalright has been a player in the premium heat sink field a long time and has a reputation of innovation, solid engineering and good quality products at fairly reasonable prices. Since they are a leading name in high performance aftermarket cooling, they have pretty much been the target to beat for other companies to strive for, especially since they introduced the original TRUE in 2007. The TRUE has proven to be a top performing cooler since it’s introduction, but had a quality control issue with base finishes for the first few years. Unless you were very lucky, you would end up having to lap the base due to the poor finish on the base not giving optimum contact with the IHS on the processors, leaving some cooling performance on the table. It seems to me though that Thermalright has addressed these problems starting with the TRUE Revision C and they don’t need to be lapped any more for optimum performance. Of the four heat sinks I am testing above from Thermalright, the only one that came with a poor base finish was the Copper TRUE, which was manufactured before they released the Revision C models and had to be lapped. ProlimaTech is a relative newcomer to the high performance air cooling scene, but has shown up with some real winners so far. Their first product released is the Megahalems, which I will be testing here. This is an original model Megahalems, but is functionally the same as the later Revision B models that are being sold now. Finally we are looking at the Noctua NH-D14. Noctua is an Austrian based company that started being sold here in the USA a few years ago and have shown themselves to produce some great cooling products. And the NH-D14 is a very large heat pipe solution that is a bit unlike the other heat sinks in this test in that it comes with a 120 x 25 mm fan and a 140 x 25 mm fan and is designed to work with both installed. We will see if the NH-D14 can deliver the goods and cool as effectively as the other choices here.

Thermalright TRUE Copper

First up to be looked at is the oldest model here, the TRUE Copper. It is an evolutionary offshoot of the original TRUE and is simply the original TRUE, but done completely in copper. This also makes it the heaviest heat sink of the roundup by far at almost 2 kilograms and requires extra care when moving a system around with it installed. This cooler came with 4 fan clips to mount 2 fans in a push-pull configuration and the older “X” style mounts for both LGA775 and LGA1366, as well as the original style AM2 mount. The specs for it are as follows:

• Dimensions: L 132 x W 63 x H 160.5 mm
• Weight :1900g (Heat sink Only) (manufacturer)
• Weight: Measured weights as follows: 1896.5 grams (Heat sink only), 1902.7 grams with 4 fan clips
• Recommended Fan (120 x 25 mm)
• Heat pipes: six Heat pipes
• Material : Copper
• 52 stacked fins

I bought this heat sink new right after they were introduced to the market. As delivered, the general finish on the heat sink was good and came with mounting hardware for the Intel and AMD sockets in current usage at the time. However, the base finish on it was absolutely horrible. It looked as though they had tried to machine a convex base finish on it, but it was off-center and gave much worse performance than the regular TRUE heat sink I had at the time. After lapping the base, the performance of it improved markedly, with it slightly outperforming the original TRUE model. This is the only heat sink in this review that has had anything done to the manufacturer’s base finish.

Thermalright TRUE Rev. C 1366RT

The next heat sink to look at will be the TRUE Revision C 1366RT. This heat sink has been revised from the original TRUE configuration by having a “V” cut into the center of the fins from the bottom going up to almost the top fin and Thermalright also changed the fin count and spacing from the original model. This is the retail model for LGA1366 mounting only and came with a fan holder and Thermalright branded fan instead of the wire clips normally associated with the TRUE series. It also included the new Revision B mounting system, which is much more secure than the older “X” style mounting system. The specs for it are as follows:

Heat sink

• Dimensions: L 132 x W 63 x H 160.5 mm
• Weight: 790g (Heat sink only) (manufacturer)
• Weight: Measured weights as follows: 731 grams (Heat sink only), 966 grams complete
• Heat pipes: Six heat pipes
• 48 stacked fins

Fan

• Dimension: L 120 x W 120 x H 25 mm
• Bearing type: FDB (Fluid Dynamic Bearing)
• Rated Speed: 1600 RPM
• Noise Level: 28.0 dBA
• Measured weight: 171.2 grams
• Air Flow: 63.7 CFM

This heat sink was bought by myself earlier this year, right around the time the Thermalright Venomous X was introduced in order to see if there was any discernible differences between the Rev. C and original TRUE models and also to compare it with the Venomous X. The biggest difference I found with it was the inclusion of the newer Rev. B mounting hardware, which I find to be a substantial upgrade compared to the original “X” type mounting hardware included with my Copper TRUE for LGA1366. While the original “X” type clamping system can be made to work well, you generally had to do some mods to it for optimum clamping and holding pressure. The Rev B mount does away with the need for mods to get a secure and tight mount. Also with this retail model TRUE was the inclusion of a plastic fan holder instead of using wire clips to mount the fan like the earlier TRUE models used. But, the retail version also didn’t include a way to mount a push-pull dual fan setup on it without having to buy some fan clips or another plastic fan holder either. The base on this heat sink is bowed with it being bowed across the long axis, point loading the center of the IHS under compression. Like modern bowed water blocks, this helps optimize heat transfer between the center of the IHS and heat sink base.

Thermalright Venomous X

Next we will look at Thermalright’s next evolution of TRUE heat sink design, the Venomous X. It is more of an evolution of the TRUE, in my opinion than a totally new model. They changed the fin shape of the heat sink from the TRUE, being a bit thicker through the middle of it rather than a bow type look to the fins, but it is readily apparent where it’s design originated from. Also, Thermalright also changed the mounting hardware for Intel based systems, giving you a way to easily change the mounting pressure applied to the heat sink base and also giving you a mounting base that works across LGA775, LGA1156 and LGA1366 platforms with one set of hardware. The specifications for the Venomous X are as follows:

• Dimensions: Length 127 mm x Width 63 mm x Height 160 mm
• Weight: 755 g (excluding fan and bracket system) (manufacturer)
• Weight: Measured weights were as follows: 757 grams (Heat sink only), 763 grams with 4 fan clips
• Heat pipes: 6mm sintered heat pipe x 6 units
• Copper base: C1100 pure nickel plated copper base, with ultra-shine mirrored surface.

As stated above, the mounting system for the Venomous X was changed from even the Rev. B model clamping system used by the late model TRUE heat sinks, but the 2 mounts are fully interchangeable between the 2 different series heat sinks. The Intel mounting system included with my Venomous X is designed to work with LGA775, LGA1156 and LGA1366 sockets with the same hardware. Another change with the Venomous X is the wire clips that are used to secure the fans to the heat sink. There are holes drilled into the fins at the top and bottom of the heatsink for the clips to slip into and stay semi-permanently in place, compared to the wire clips hooking on the side like the TRUE series were designed to do. This is a welcome change in my opinion, as I find it makes it a bit easier to swap out fans since the clips simply pivot out of the way and stay on the heat sink. Also with this mounting system, Thermalright also gave us a way to vary the mounting pressure in the form of a knurled screw that goes through the middle of the top mount and nestles into the indentation in the top of the heat sink base and lets you vary the clamping pressure between 40 to 70 lbs of clamping force. By the way, this mounting system also works just as well on the TRUE series and the True Spirit as it does on the Venomous X. If Thermalright would decide to sell this mount separately in the future, it would be a viable choice for buying to use with your TRUE or True Spirit, especially since it works on the 3 different Intel sockets in use on single processor systems. The base on this heat sink is bowed with it being bowed across the long axis, point loading the center of the IHS under compression.

Cogage True Spirit

The final Thermalright product we will look at actually isn’t marketed by Thermalright itself, but comes from their subsidiary company, Cogage. This heat sink is named the Cogage True Spirit and it’s design looks to be a derivative of the Ultra 120/TRUE design. It only uses 4 heat pipes instead of the 6 heat pipes used in the other three Thermalright products. It also is by far the most inexpensive of all the heat sinks being tested here, selling for $35.99 at CrazyPC and this includes a PWM fan with it. The model I am reviewing is the newest revision of the True Spirit, which uses the old “X” style Thermalright retention bracket and backplate that came with the older models of TRUE instead of the push-pin type mounting system that this heat sink was originally introduced with. Here are the specifications for it:

Heat sink

• Dimensions: L 133 x W 53.44 x H 163.8 mm
• Weight: 680g complete (manufacturer)
• Weight: Measured weights as follows: 537 grams (Heat sink only), 637 grams complete
• Heat pipes: Four heat pipes
• 48 stacked fins

Fan

• Dimension: L 120 x W 120 x H 25 mm
• Rated Speed: 1000-1500 RPM
• Noise Level: 19.6 – 37.4 dBA
• Air Flow: 35 – 66.5 CFM

As stated above, this heat sink includes the older “X” style of LGA1366 clamping system in the box. The original release of the True Spirit used the “X” style clamp without a backing plate, but used the cursed Intel push-pins. They revised the mounting system by including the backing plate and tension spring screws that came with the TRUE heat sinks, which is a better mounting system than the push-pins, in my opinion. The “X” style top still doesn’t give optimum clamping pressure without doing a washer mod, but does work better with this heat sink than with the TRUE due to the True Spirit being a lighter heat sink. This heat sink also uses the same wire clips that are used on the Venomous X, in case you want to get an extra set of clips so that you can run a push-pull setup on this heat sink. The base on this heat sink is also bowed with it being bowed across the long axis, slightly point loading the center of the IHS under compression.

Prolimatech Megahalems

Next we come to the Prolimatech Megahalems. This was the first heatsink that Prolimatech had on the market and after I bought my example, they upgraded the looks slightly by using a stainless steel top fin instead of an aluminum top fin like the model I own, to better resist tarnishing. They also included 2 sets of wire clips with the Revision B model instead of 1 set of clips that were included in my original model as well as a mounting adapter that works across LGA775, LGA1156 and LGA1366 instead of just LGA775 and LGA1366 mounts like mine. But functionally it is exactly the same heat sink as my original model Megahalems. The specifications for the Megahalems are as follows:

• Dimensions: L 130 x W 74 x H 158.7 mm
• Weight: 790 grams (manufacturer)
• Weight: Measured weights as follows: 779.5 grams (Heat sink only), 786 grams with 4 fan clips
• Heat pipes: 6 X 6 mm heat pipes

The Megahalems impressed me with the quality aluminum mounting adapter that comes with it. Instead of using stamped steel components like Thermalright and Noctua use, they used thicker, more rigid aluminum components and make for a very good mount with hardly any side to side play in the mount. The base is also well finished, with it being slightly crowned across the short axis, giving a “bowed” finish similar to today’s high performance water blocks by point loading the center of the HIS for a better contact patch while under compression. The heat sink itself is thicker in cross section compared to the Thermalright offerings and the fins are a 2 piece stack that are separated by about a ¼” air gap between the heat pipe stacks.

Notctua NH-D14

Last but not least, we are looking at the Noctua NH-D14. This heat sink was the second most expensive heat sink in this review, but it also comes with two fans included with it. The two premium fans included with it are sized differently, with one being sized at 120 x 120 x 25 mm that acts as a pusher fan into the first fin stack and the other is a 140 x 140 x 25 mm fan that sits squarely in the middle of the heat sink between the 2 fin stacks that are connected by 6 heat pipes and blows air from the first fin stack through the second fin stack. This design is somewhat reminiscent of the earlier Thermalright IFX-14 offering and is the most massive heat sink being tested in this review. Noctua is an Austrian based company that came into the US cooling market a few years ago with some solid designs; nothing extremely innovative but very well designed and executed. This heat sink also comes as a complete package, with mounting adapters for all commonly used Intel socket designs as well as AMD AM2/AM3 mounting adapters. It also includes some low noise resistive adapters to slow the rotational speed of the included fans down also, if absolute silence is wanted over cooling efficiency. For non-stock fan testing with this heat sink, I also had to change from the Scythe S-Flex SFF21G fans to Yate Loon High speed fans because of the mounting system that Noctua uses for their fans not being compatible with the slightly thicker flange area of the S-Flex fans. Noctua uses some plastic pins that stick through the holes in the fan flange face and with the thicker face of the S-Flex, the pins didn’t stick through far enough to engage the holes in the plastic pins with the wire fan holders. The Yate Loon fans flow slightly more air than the S-Flex “G” fans, but it’s not enough to cool appreciably better. The specifications for this heat sink are as follows:

Heat sink

• Dimensions: L 130 x W 140 x H 160 mm
• Weight: 900 grams (without fan) (manufacturer)
• Weight (with fan)1070/1240 grams (1 fan/2 fans) (manufacturer)
• Weight: Measured weights as follows: 928 grams (Heat sink only), 1108 grams w/NF-P12 fan, 1091 grams w/NF-P14 fan, 1271.5 grams complete
• Fan compatibility 140 x 140 x 25 mm and 120 x 120 x 25 mm
• 6 heat pipes feeding into 2 separate fin stacks

Fans

• Models NF-P14 & NF-P12
• Dimension: L 140 x W 140 x H 25 mm & 120 x 120 x 25 mm
• Rated Speed: 1200 & 1300 RPM
• Noise Level: 19.6 & 19.8 dBA
• Air Flow: 64.9 & 54.3 CFM

Test Setup

My testbed setup is a modified older case design, but I think it represents an average case as far as cooling efficiency is concerned. The testbed setup is as follows.

• Case – Chieftech clone of the Antec 1040 case series, with the original 80 mm exhaust fans being removed and the holes enlarged to mount two 92 mm fans externally on the outside of the rear case bulkhead instead of internally.
• Motherboard – Asus P6T
• Processor – Intel Core i7 Extreme 980X, overclocked to 4000 MHz @ 1.304 v.
• Ram – Kingston HyperX T1 DDR3 2000 3X2 GB kit or Corsair XMS3 DDR3 1600 (used when clearance issues were encountered)
• Video Card – eVGA 7900GTX
• Power Supply – HEC Cougar series (German HEC, not US model) S700
• Western Digital 250 GB SATA hard drive
• Lite On DVD-RW drive
• Windows Vista Ultimate 64 Service Pack 2

Fans used for testing were as follows:

• Stock Thermalright fan included with the TRUE 1366RT for single fan testing; weight measured at 171.2 grams
• 2 – Scythe S-Flex SFF21G fans for push-pull test configurations – 1900 RPM, 75.0 CFM, 35.0 dBA; weight measured at 182.1 grams per fan
• 3 – Yate Loon D12SH-12 fans on NH-D14 only – 2200 RPM, 88.0 CFM, 40 dBA; weight measured at 137.5 grams per fan
• Cogage PWM fan (with PWM function disabled) on the True Spirit only
• Stock Noctua fans on the NH-D14 only
• Sanyo Denki SanAce 9CR1212P0G03 120 X 76 mm fan for extreme cfm testing – 6200/2700 RPM, 300 CFM, 70 dBA, 1.93 inches H2O static pressure; weight measured at 753.3 grams with 1 finger guard installed
• The Thermalright plastic fan holder which comes with the TRUE Rev. C 1366 RT heat sink can also be bought separately and it weighs in at 63.7 grams.

Since I have an Ohaus triple beam balance already, I decided to weigh each heat sink and the add-on fans in order to get an idea of how accurate the weights quoted by the manufacturers really are. I also weighed the fans and fan holders (in Thermalright’s case) so that you could get an idea of how much weight is being suspended above the motherboard when the heat sink is mounted. I didn’t weigh the mounting systems as I feel that weight doesn’t factor into the potential bending moment if the case were dropped or jarred hard. There is the potential when a case is handled roughly that the motherboard would be damaged due to the weight of the heat sink and these taller heat sinks can exacerbate the situation because of their tallness and weight mounted high above the motherboard giving the heat sink a high bending and cross-load moment when dropped or jarred. For temperature monitoring purposes, I am using Real Temp 3.40, with logging enabled at 2 second intervals. For outside air temperature monitoring I used a digital thermometer and kept the temperature in the room between 20.5 to 21.1 °C (69-70 °F), measured at the front of the case. If room temperatures exceeded these parameters I re-ran that individual test run. Hopefully for future reviews I will be able to upgrade to a logging digital monitor that can log the intake temperatures so that I can more accurately measure delta T. For loading the CPU, I used Prime95 version 25.8 using in-place large FFT’s and ran it for 30 minutes to stabilize temps. After 30 minutes I would exit Prime95 and let the CPU idle for 5 minutes. The highest recorded temperature from the hottest core for each run was then recorded off of the Real Temp log, the lowest temperature on any core was recorded and the average temperature on the hottest core was calculated during the load portion of each run. I tested each heat sink in the following configurations. For heat sinks that included a fan in the retail package, I tested with the included fans. I tested the heat sinks except the NH-D14 using the Thermalright fan included with the TRUE Rev. C 1366 RT in push configuration only. It has exactly the same specs as the Scythe S-Flex SFF21F fan and I imagine that Thermalright sources it from the same OEM as Scythe. For push-pull tests, I tested all the heat sinks except the NH-D14 with two Scythe S-Flex SFF21G fans. For the push-pull testing on the NH-D14 I tested it with 3 Yate Loon D12SH-12 fans since I couldn’t get the fan clipping mechanism that Noctua uses to work with the Scythe fans. For clipping on the third fan onto the NH-D14, I used one set of the fan clips for 25 mm thick fans that comes with the Prolimatech Megahalems. Those fan clips work well on the NH-D14 to hold a 25 mm fan on either end of the NH-D14. Each heat sink was mounted on the system 3 times and the best run of each was taken for this review. Finally, for extreme CFM testing (for benching on air) I tested each heat sink with a fan I have dubbed The Beast. This Sanyo Denki compound fan is the most powerful 120 mm form factor 12 volt I know of and nobody in their right mind would use this fan for everyday usage as the noise is bloody awful. But it is a great fan for bench testing on high end air and testing with this fan will give us an idea of whether these heat sinks can improve their efficiency at removing heat with extreme airflow. For extreme CFM testing on the NH-D14, I gutted a fan out and mounted the frame between the heat pipe stacks to help guide the air through the second heat pipe stack for more efficient cooling. I was able to use the stock mounting systems on the Thermalright and Cogage heat sinks to mount this fan, but with the Megahalems and the NH-D14 I had to use zip ties to securely mount the fan to the heat sinks. Also during this phase of testing, the case side was removed to keep from recycling hot air and relieve the pressure that The Beast puts on the exhaust fans on the case.

The Results

Now finally on to the results. First up are the runs done with the retail fans that came with the heatsink on the heatsinks that came as a complete cooling solution. The NH-D14 was tested in it’s stock configuration with both the stock fans installed, since that is the way it comes from Noctua. The True Spirit was tested with it’s stock fan with the PWM function turned off in bios, to let it spin at full speed. And the Thermalright TRUE Rev. C 1366RT was tested with the included Thermalright fan.

Next are the runs done with a single Thermalright TR-FDB-1600 fan. The NH-D14 was tested in it’s stock configuration with both the stock fans installed, since that is the way it comes from Noctua and it’s made to run with 2 fans. The data from the Noctua and the TRUE Rev. C 1366RT are a repeat of the above data.

Now we come to the high performance testing. For all heat sinks except the Noctua, the Scythe S-Flex SFF21G fans were set up in a push-pull configuration. For the Noctua, there were 3 Yate Loon high speed fans set up with one as an intake, the second in between the heat pipe stacks and the third used as an exhaust helper on the second heat pipe stack.

Finally, we come to the extreme CFM test on these heat sinks. All were tested with the Sanyo Denki set up in a push configuration. The Noctua had a fan frame inserted between the first and second heat pipe stack to funnel and guide the air through the second heat pipe stack.

Conclusion

There are no losers in this group of heat sinks. All do a superb job of keeping an overclocked Gulftown from broiling itself and are well manufactured. All have good to excellent mounting systems and had no clearance problems with any components around the socket. The NH-D14 does have issues with memory that uses tall heat spreaders but this can be worked around by using ram with short or removable heat spreaders or simply by mounting the intake fan offset a bit to clear the ram. But with that said, in normal, out of the box realm of operation, the good old tried and trusted TRUE Rev. C and the Noctua NH-D14 ran head to head and are a toss-up for the best out-of-box cooler. The Noctua showed a slightly lower average temp on the hottest core on this processor while under load, but the TRUE-C had the lower absolute max temp. The Venomous X comes in a solid third, then the True Spirit and TRUE copper run neck and neck for fourth and fifth place. Finally, the Megahalems trails the rest a bit, but still is a very good choice for cooling your hot running i7 system. It’s average in this configuration was only around 4 ½° C higher on average load temp than the NH-D14. In the high performance setup, the Noctua NH-D14 ended up with just an ever so small performance edge over the TRUE Rev. C. Again, both were neck and neck on the high temp recorded but the NH-D14 did record a 1° C lower idle temp and it’s average load temp was almost ¾° better than the TRUE. Again, the Venomous X came in a solid third (with thoughts of making a run at second). In fourth place the venerable old TRUE copper showed its good looking head while the True Spirit was closely following it in a solid fifth place. Again, the Megahalems is trailing the pack, but it’s only lagging the NH-D14 by around 2 ½ °C. In the extreme performance category we see a bit of a surprise here. One of the leading heat sinks in the first 3 configurations kind of falls flat on its face. The NH-D14 didn’t see nearly the performance improvement that the rest of the heat sinks did with extreme cfm and static pressure. The leader in this configuration was the TRUE Rev. C, with a solid 1 ¼ degree lead in average load temp over second place. In a bit of a surprise, the cheapest heat sink of this roundup comes in second in the extreme CFM test, the Cogage True Spirit. Closely following it was the Venomous X, followed by the TRUE copper and the Megahalems. Sucking up in last place this time was the Noctua NH-D14, with an average load temp 4 1/3 degrees higher than the leading TRUE Rev. C. It looks like its configuration just doesn’t respond as well to extreme flow rates and static pressure than the other heat sinks.

So who is the overall winner? I’m just going to leave that for you to decide. All have their merits and all are excellent choices for cooling. The old coot of the bunch, the TRUE Copper didn’t perform at the lead of the pack, but is a very good looking and kind of rare heat sink. The Cogage True Spirit is solidly in first place in the bang/buck category. The Venomous X is a very good performing heat sink with a mounting system that is versatile for mounting across all Intel platforms and is also tunable in mounting pressure. The Megahalems has the absolute best mounting system in my opinion and the aluminum hardware in their mounting system just screams “quality”. The NH-D14 comes in and shows us that the Austrians have some excellent engineering and quality control and that they can produce a cooler that is a pack leader. And while it is a bit more expensive than every other cooler in this roundup except for the TRUE copper, it also includes 2 high quality fans that promise you years of trouble free operation. Finally, we come to the TRUE Rev. C. Although it’s initial design is several years old, Thermalright reworked the initial design by giving slightly wider fin spacing than the original as well as redesigning the interior of the fins a bit, putting a “V” shaped cut into the fins in the middle of the stack. It’s redesign was effective too, with this heatsink performing right up there with the NH-D14 in normal and high performance modes and absolutely covering the field in the extreme performance category. They also re-engineered the mounting bracket from the original design for LGA1366, which gives a much more stable mount without having to mod the mounting system. In closing, I hope that you have liked my review and hope that this will help you out in your cooling decisions. And I would also like to thank a couple of vendors that have helped me out by giving me some price breaks while acquiring the heat sinks and fans I use for testing. Jab-tech and Sidewinder Computers are both quality places to shop your cooling needs, with competitive pricing and fantastic customer service. I thank you guys for helping me out; it is much appreciated. And to you readers out there; I have enjoyed putting this review together for you and hopefully I will be able to do some more reviews in the future.

– muddocktor