Introduction

As we near the end of the year I’m happy to present you with my latest roundup of Athlon 64 heatsinks.



What I have in store for you:

Arctic Cooling Freezer 64 Pro *

Noctua NH-U 12 *

Noctua NH-U 9 *

Sharkoon Red Shock *

Silverstone NT02 *

Scythe Samurai Z *

And from our previous HSF reviews at Madshrimps I have included these for reference and comparison:

AMD Stock Cooling (from 3200+)

Arctic Cooling Freezer 64 – very good price/performance unit

Scythe Shogun – top performance at low noise

Thermalright XP-120 – all time favorite of many

Zalman CNPS9500CU – latest in Zalman’s popular CNPS series

If you want to know more about the heatsinks mentioned in the list above please visit our previous roundups:

AMD A64 Heatsink roundup Summer 2005

AMD A64 Heatsink roundup Q2 2005

Intel P4 Heatsink Roundup Q1 2005

Intel P4 Heatsink Roundup Q4 2004

This year’s last roundup took several weeks and many hours to complete; the main reason for this is because I decided to test the heatsinks in a case as compared to my earlier roundups where I just let the motherboard sit on top of a cardboard box.

Test Setups and Methodology

JMke's Test Setup
CPU	A64 3200+
Mainboard	DFI NF3 Lanparty
Memory	1 * 256Mb PC3700 OCZ
Other	ATI R9000 Passive Cooling Chaintech FX5900XT with Arctic Cooling Silencer Silverstone EFN-300 300W Passive Cooled PSU Antec SmartBlue SL350P 350W Active Cooled PSU Maxtor 120GB IDE HDD

in-take temperature was measured at 22°C for all tests, but temp fluctuations, different mounting and user error can account up to 1-2°C of inaccuracy in the obtained results. Please keep this in mind when looking at the results. Each heatsink was tested repeatedly; if we got questionable results the test was restarted.

Noise level of each HSF combo was recorded with SmartSensor SL4001A, the sensor was placed ~50cm away from the case. The lowest dBA reading in the test room was 32.5dBA with everything turned off!

System was stressed by running K7 CPU Burn for 30min (after Thermal Compound’s burn-in); this application pushes the temperature higher then any other application or game we’ve yet encountered. Speedfan was used to log maximum obtained temperatures.

Arctic Silver kindly send us their “Lumière” thermal testing compound which has the same colour as Ceramique, but only a break in time of 30min!

Arctic Silver’s ArctiClean was used to clean off thermal paste of the CPU and heatsink between tests

Fans used for comparison

Delta NFB0912L 92mm: 42CFM

Vantec Tornado 92mm: 119CFM

Papst 120mm 4412 F/2GLL: 40CFM

Titan 120mm: 115CFM


The Case

In the past I always worked in a case-less setup, which allowed easy installation and removal of the heatsinks, the results obtained were valid when compared between the different heatsinks included in that particular roundup. However with the introduction of different Tower-like heatsinks the orientation of these heatsinks inside a case has been giving them a noticeable advantage over the conventional “fan blowing down onto the heatsink/motherboard” design.

So for this roundup I decided to use a case, not just a randomly selected one though, it has to have room for rear 120mm fan, as this offers the best cooling/noise balance. With the industry’s continuing obsession for silence the Antec Sonata II will make a perfect host for my test setup. It comes with 120mm rear fan which you can run at 5/7/12v by use of a small switch.

Since I’m only using an Athlon 3200+ for my tests, it would be interesting to overclock the CPU so its maximum heat output increases and it can simulate a higher clocked Athlon 64. At default it’s rated at 89W, increasing clock speed to 2400Mhz with 1.7v vcore it’s putting out ~125W according to this little calculator. In my days of Athlon XP HSF testing an increase of 0.1v vcore would result in 4-6°C higher CPU temps, so without much surprise the temperature results here with the 1.7v Athlon 64 were much higher.

To cater to overclockers as well as people who prefer silence I decided to test with 2 different power supplies.

1) Actively cooled Antec SmartBlue SL350P 350W which features 1x80mm outtake and 1x92mm intake.
2) Passively cooled Silverstone EFN-300 which can run quite hot

Noise was recorded approx. 50cm away from the case at an angle, here’s a (very bad) drawing of how the dBA meter was position opposite the case and the test-room.


Green box = Sonata II – White Dot = dBA meter

3 different test scenarios were configured as such:

Test Setup 1: Overclocked and Silent



- Antec Sonata II + Silverstone Passive PSU
- AcoustiFan DustPROOF 120mm @ 5v in the rear as outtake (mounted with Acousti Fan Gasket)
- Athlon 64 3200+ @ 2400 – 1.7v vcore
- nVidia TNT2 Passive cooled video card
- Noise produced with system running without HSF fan: 34dBA @ 50cm

Test Setup 2: Overclocked and Noisy



- Antec Sonata II + Antec SmartBlue PSU
- AcoustiFan DustPROOF 120mm @ 7.5v in the rear as outtake (mounted with Acousti Fan Gasket)
- Antec Sonata II’s 120mm fan in front @ 7v (mounted with Acousti Fan-Mounts)
- Athlon 64 3200+ @ 2400 – 1.7v vcore
- Chaintech FX5900XT with Arctic Cooling Silencer
- Noise produced with system running without HSF fan: 50dBA @ 50cm

Test Setup 3: Not Overclocked and Noisy
- Antec Sonata II + Antec SmartBlue PSU
- AcoustiFan DustPROOF 120mm @ 7.5v in the rear as outtake (mounted with Acousti Fan Gasket)
- Antec Sonata II’s 120mm fan in front @ 7v (mounted with Acousti Fan-Mounts)
- Athlon 64 3200+ @ 2200 – 1.5v vcore
- Chaintech FX5900XT with Arctic Cooling Silencer
- Noise produced with system running without HSF fan: 50dBA @ 50cm

50dBA with active PSU, video card and more case cooling is quite loud, the main culprit is the Antec Smartblue power supply which ramps up its output 80mm fan when the case temperature goes up.

What was measured?

The CPU temperature was measured with SpeedFan and highest value recorded

Temperature of air coming into to the case at the front

PWM temperature through SpeedFan, this represent the area around the CPU socket, the power management caps which you see on a motherboard, they are there to make sure the power which is fed into the motherboard coming from the PSU is filtered and delivered the CPU and other components. Too high temperature will cause Vcore fluctuations which in turn causes system instability.

The VGA temperature (only measured in Test #2 & #3 with FX5900XT onboard probe)


Let's get started with our first new contestant ->

Arctic Cooling Freezer 64 Pro:
Supplied by: Arctic Cooling

Arctic Cooling introduces an improved Freezer 64, the Pro version adds a larger fan, more heatpipes and it’s a bit larger then its predecessor.

It can be found priced at €15~20!



Specifications :

Compatibility:

AMD: Athlon64/64+ Socket754/940/939

Fan included: 100mm custom design
Fan Speed: 2200rpm
Air Flow: 40CFM

Freezer 64
Heatsink + Fan Dimensions: 92x114x120mm
Weight: 460gram

Freezer 64 Pro
Heatsink + Fan Dimensions: 107x96.5x126.5mm
Weight: 528gram

In the Box :

A case badge

Step by step manual

there’s no tube of thermal paste included but their MX-1 comes pre-applied on the base


Construction :

The Freezer 64 Pro resembles the “none Pro” version closely, but the larger width and 3rd heat pipe give it away. The 3 U formed heatpipes are joined in a copper base, a series of aluminum fins dissipate the heat, the bottom rows of fins are further spread out to allow air to pass more easily, as the other side of these fins is bend downwards to guide air towards components around the CPU socket.



The base comes with pre-applied MX-1 paste which performs on par with the best out there. And is actually the best mating partner of the base.



As you can see in the photo below there are grooves from the CNC machine visible, the MX-1 paste allows for these grooves to fill up effectively and thus guarantee good thermal performance. I’ve swapped MX-1 and Arctic Lumiere but could not found any difference after 1 day of stress testing, the MX-1 however needs several days to “settle”, so in the end, you’ll be better of the with pre-applied thermal paste Arctic Cooling supplies.



This is what Magnus Huber from AC has to say about the base and the MX-1 paste:

Important is that grease and surface roughness should correspond. We guarantee an absolutely flat surface by a CNC cut. This leaves grooves on the surface. For the thermal paste there is a basic rule: the higher the viscosity, the higher the thermal conductivity. The problem, higher viscosity also makes it more difficult to squeeze the grease and make sure that there is no air in-between the cooler and the CPU anymore. The film thickness of a viscous paste is therefore higher and this lowers the performance. Our job is it now to achieve a thin paste film with a very viscous thermal compound. Grooves help to do this job. So our grooved cooler surface works perfectly with a thermal compound with high viscosity and bad with a compound with low viscosity. If we would make a mirrored surface, it would work better in combination with a thermal compound with low viscosity.

The fan mounting method is one the best I’ve encountered, the plastic fan bracket clips on to the aluminum fins, the fan itself is decoupled from the frame by an ingenious system which effectively reduces vibration noises. The fan cable is sleeved and sufficiency long.



Installation next ->

Installation :

The Freezer 64 Pro uses the stock AMD bracket, which makes installation a breeze, removing the fan during installation gives you plenty of working space and you can easily clip it back on afterwards. The drawback of this custom fan bracket is loss of compatibility, as no other fan can be installed onto this heatsink, lucky for us Arctic Cooling’s 100mm fan is of high quality and won’t disappoint.

The clip works the same as with the Stock AMD heatsinks, but you can change the heatsink’s orientation, by removing the small screw you see in the photo below, you can install the heatsink so it blows air towards the top or rear.




The normal Freezer 64 does not offer this possibility and with the DFI board I’m using this heatsink ended up blowing air upwards:



Luckily with the Freezer 64 Pro you have a choice:





Performance and Noise :

Let’s see how it compares to the stock A64 cooling and its predecessor the Freezer 64



In the first test configuration the temperatures run up very high (which you will see will be the trend with this config). At 100% fan speed the Pro is quite noisy but is also able to keep the overclocked CPU under 60°C. Slowing down the fan shows an increase of 7°C, noise drops to a whisper quiet 35dBA, you’ll have to visually check to know if the fan is running, because it’s dead silent. The PWM area is running hot as there is not a lot of air circulation; the downward bend fins on the Freezer 64 Pro prove to be effective, while with the “none-Pro” version PWM reaches a worrying 80°C!

Time to add more case cooling:



Again both coolers are tied at 100% fan speed; only at 50% the Pro version is able to take a small lead. The PSU is drowning out the CPU fan noise completely and at 50dBA the system is clearly audible in a quiet room.

Putting the CPU at stock speeds gives this:



Both Freezer 64 coolers perform identical at 100%, at 50% the Pro has the edge. The Freezer 64 is aligned to blow air upwards to the PSU instead of the rear due to the alignment of the socket on this DFI board. To level the playing field I installed the Pro to also blow towards the top:



Although the difference is not big, it’s recommended to let the heatsink blow heat out towards the rear, which makes sense of course.


The Verdict:

The Pro version performs on par with the Freezer 64 when running at 100% fan speed in my testing, at which point the Pro is louder, but when the fan speed is reduced the larger Pro heatsink shows it muscles, keeping the CPU cooler as well as the surrounding components.

Featuring easy installation, excellent out of the box performance and a very competitive price the new Arctic Cooling Freezer 64 Pro will be very popular with good reason!


PRO
Excellent performance
Quiet operation when fan is slowed down to 50%
Easy Clip-On Installation
No matter the CPU socket orientation you can install the heatsink to blow air towards the rear.
Pre-applied MX-1 high performance thermal paste
Very competitively priced

CON
Does not offer a performance improvement over the original Freezer 64 in some cases
Custom fan bracket does not allow other fans to be installed
Quite noisy with fan at full speed (44dBA @ 50cm)

Where to buy?


onto our next victim ->

Scythe Samurai Z:
Supplied by: Scythe

Scythe’s Samurai series of heatsinks has been with us for some time now, we reviewed one (SCSM-1000) back in May of 2004, the new version comes with a “Z” tag, 92mm fan and heat pipes joined in a copper base.

While I couldn’t find a price on the latest one, the previous SCSM-2000 (which is a revision of the SCSM-1000) can be had for €25-30, it’s half the price of Scythe high performance “Shogun”.




Specifications :

Compatibility:

AMD: Athlon64/64+ Socket754/940/939

Intel Socket 478

Intel Socket LGA775

Fan included: 92x92x25mm
Fan Speed: 2000rpm
Air Flow: 32.4CFM

Heatsink + Fan Dimensions: 128x85x98mm
Weight: 360gram

In the Box :

Step by step manual for each platform

mounting clips

a small amount of generic thermal paste


Construction :

The Samurai Z is made from a combination of two materials: 2 copper heat pipes run through the heatsink and are joined in copper base , small aluminum fins help dissipate the heat.



A 92mm fan from Scythe (model AD912LS-A76GL) is mounted to blow fresh down towards the base; the clips provided are not the easiest ones to remove/install but get the job done. Bridged or webbed fans are incompatible with these clips however (example)



A view from the top (above) and side (below)



The base is very smooth, close to a mirror finish; I would almost think that it was lapped manually. No machine marks are visible.



Close-up of the base:





Installation next ->

Installation :

This heatsink can be installed on all current Intel and AMD platforms like most of Scythe previous CPU cooling products, so I was expecting plenty of mounting hardware pieces, much to my surprise all I needed for installation on AMD Socket 754/939/940 is this:



They clip on at each side of the heatsink and with them installation was extremely easy, quite an improvement over the previous methods used by Scythe which consisted of screws, a back plate and even more screws.




Performance and Noise :

Let’s see how it does compared to the stock A64 cooler



First thing which I noticed was the quality of the included 92mm fan; it performs better than my reference Delta fan which is already quite high up the performance/noise ladder.

The temperature of CPU shows an improvement of almost 5°C, and this at a lower noise level, the dBA meter displayed a healthy 10dBA drop.



For the Overclocked and Noisy test I added the Vantec 92mm fan in the mix, and as you can see there’s a reason why it should never be considered for any “silent computing”. 71dBA at 50cm from the case it’s deafening. Coupled to the Samurai Z it offers excellent performance (no surprise); the other results all show similar drops in temperature compared to Setup1, this is of course thanks to the extra case ventilation.



At stock speeds I was able to reduce the fan speed of the Delta and Stock fan to 50%, this caused a system crash with the previous test setups, but now the CPU is running a lot cooler. But this reduction in fan speed is sadly to no avail as the PSU is too loud and the drop in noise can not be noticed.

In my sound tests with the Samurai Z in setup1 I recorded a 34dBA noise level with the stock fan running at 50%; If you are not overclocking your CPU and have a quiet PSU this fan/heatsink combo will perform well and do it silently.



The Verdict:

The new Samurai Z is surely an improvement over its predecessor, the silent operation (even with the fan running @ 100%) combined with easy installation will make it an interesting choice for those looking for an all-around performer. While in my tests it does not come out on top performance wise, its plug and play installation for all platforms is a welcome and unique asset.

PRO
Included 92mm fan offers excellent noise/performance ratio
Extremely easy installation
Offers decent cooling
Compatible with all current AMD and Intel sockets
Fair price

CON
fan mounting clips don’t work with bridged fans

Where to buy?

Sharkoon Red Shock:
Supplied by: Sharkoon

Sharkoon has been expanding their product line up for quite some time know, they made a good impression on me with tower AXP heatsink back in 2004. Today it’s different design they send me, a fully copper, rather compact heatsink which does not come with its own fan, but allows you to mount 80/92/120mm of your choice. Its price point is set near €45-50.

On the product page at the Sharkoon website this is what they have to say about the Red Shock:



Let’s find out if it ready for silent computing.



Specifications :

Compatibility:

AMD: Athlon64/64+ Socket754/940/939

Intel Socket 478

Intel Socket LGA775

No Fan included but supports 80/92/120mm.

Weight: 820gram

In the Box :

Mounting hardware for each supported platform

a manual

tube of what looks to be generic thermal goop


Construction :

The construction of this heatsink is very similar to the well know Thermalright SP-94. A series of fine copper fins, and 4 heat pipes which run through them and are joined in the base.




The underside reveals holes which are used for attaching the different mounting brackets.



The base is quite smooth, with visible machine lapping threads.




Installation next ->

Installation :

This heatsink will require motherboard removal for proper installation, no matter what platform you are using; this hefty copper heatsinks needs a strong back plate for obvious reasons. After attaching the K8 mounting plates it looks like this:



The fan mounting supports 80/92mm from the get go, when you want to use a 120mm fan you have to install the addition stand offs



On the DFI board it was a tight fit as the heat pipes are sticking out and they are touching some components around the socket; this same issue I experienced also with SP-94 on different Intel boards, so far they look similar and have the same installation concerns.



Screwed on tight the Red Shock won’t come of your motherboard “by accident”, although it weighs a lot I never had a feeling this was going to be a problem for mounting. The mounting method is tried and tested and provides for a “safe fit”. In the photo below the Red Shock heatsink is installed and you can see a slight bending of one of the caps on the motherboard.



The fan bracket clips on to the Red Shock easily and removal is straight forward also,





Performance and Noise :

I set of with a series of tests using different fans at different speeds, but it soon became clear that the ideal “mating partner” for this heatsink is not the 120mm fan. The Red Shock could not keep the CPU under safe temperature limits with the Papst 120mm.



As you can see from the results above, the Red Shock performs worse than the stock A64 cooling in the silent system test. Both CPU and PWM are dangerously high. Let’s see what happens when we add more case cooling:



With the Delta 92mm fan the Red Shock is now better than the stock HSF, but the heatsink really shines when paired up the high speed Vantec 92mm fan. When you want extreme performance at extreme noise levels the Red Shock delivers!



The last test confirms my previous results.


The Verdict:

The Red Shock’s full copper construction doesn’t make it a budget solution, and since it’s been sold without a fan, there’s an extra cost involved. It performed pretty much as I expected, a compact copper heatpipe powered heatsink that shines when paired with high output fans, but when tested with silent fans fails to impress, as the performance is on par with the stock A64 HSF which you get for free with your processor. Only consider this heatsink if you are looking for extreme performance and don’t care about noise.

PRO
Compatible with all current AMD and Intel sockets
Excellent performance with high output fans
Easy fan mounting and support for 80/92/120 sizes

CON
Disappointing performance with silent fans
Motherboard removal required for installation

Where to buy?

Noctua NH-U 9:
Supplied by: Noctua

A few weeks ago I received an email from an Austrian company I had not heard of before; in it they made quite a bold statement:


Now any company can come forward with claims like this, but I was intrigued nonetheless and replied to their email. After visiting the product pages of the proposed review units (NH-U 9 / NH-U 12) I was looking forward to find out how they perform as the design looks promising.

Let’s start of with the “little” brother, the NH-U 9, as the name might suggest, is build to fit a 92mm fan. The NH-U units are “tower” heatsinks and make use of heat pipes to transfer heat from the base to the fins. Target price at TomsHW Preis-Engine is ~€37




Specifications :

Compatibility:

AMD: Athlon64/64+ Socket754/940/939

Intel Socket 478

Intel Socket LGA775

No Fan is included but supports 80/92mm

Heatsink + Fan Dimensions: 128x95x95mm
Weight: 570gram

In the Box :

Mounting hardware for each supported platform

A detailed manual

A tube of generic thermal goop

please note that the fan in the picture above is NOT included with this heatsink

For this review Noctua send along a new Coolink 92mm SWiF-922 for testing, more info on these ex-“Noiseblocker” fans here.

Construction :

The NH-U 9 is what I would call a “tower heatsink” as it towers over the other components on your motherboard; Fitted with 4 U-shaped heatpipes which are joined in a copper base, a series of aluminum fins dissipate the heat.



The base of the unit features holes for installation of different mounting brackets



The machine lapping traces are very hard to spot on the base, which is smooth.



Installation next ->

Installation :

The NH-U 9 uses a custom mounting bracket, so you need to remove the stock AMD one, and thus remove your motherboard from the case. The Athlon 64 mounting brackets installed



It turns out that the orientation of the NH-U 9 is not ideal when combined with this DFI motherboard, blowing hot air upwards. Unfortunately there is no way to rotate the heatsink towards the rear.



Fan mounting is through the use of 2 clips, installation is easy when you still have the motherboard outside the case. Once inside the case swapping fans can be a challenge, as this large heatsink will occupy the space needed to maneuver to get to the clips. Bridged or webbed fans are incompatible with these clips (example).

Noctua provides 4 strips of rubber to reduce fan vibration (so you’ll have 2 spare ones). The photo below is from the NH-U 12, which uses the same mounting method:



And here is the NH-U 9 ready to go!




Performance and Noise :

Now for the fun part, testing the performance of this unit, since they claim to be the best I decided to compare the NH-U 9 results with Thermalright’s well known XP-120 in the graphs below. (The NH-12 will face the Scythe Shogun on the next pages). When looking at the performance results you’ll have to keep in mind that this heatsink is not orientated ideally, and you might see better performance when you can blow hot air towards the rear.



The NH-U 9 pulls it off and offers a better performance/noise ratio than the XP-120, what’s more, the Coolink 92mm fan proves to be better than my reference Delta! Promising results so far, now let’s add more case cooling:


(please note that the Titan fan pushes 110CFM at 12v and doesn’t do this silently)

Again impressive results from the NH-U 9 unit in this test, although its lead over the XP-120 is diminishing. When paired up with the Vantec Tornado its offers excellent performance (to no surprise);



On par with the XP-120 with both noisy and silent fans!


The Verdict:

Once every so often a new heatsink manufacturer comes along and takes the performance crown from the current crop of contenders; Noctua certainly is aiming to be that company, and with the NH-U 9 they have a winner on their hands as it offers excellent performance, be it with silent or noisy fans. With multiple socket compatibility and competitively priced it comes recommended for those seeking a budget-wise CPU cooling solution and possibility to use a fan of their choice.


PRO
Compatible with all current AMD and Intel sockets
Excellent performance with both silent / powerful fans
Competitively priced

CON
Motherboard removal required
Can not change heatsink orientation

Where to buy?

Noctua NH-U 12:
Supplied by: Noctua

The NH-U 12 is currently Noctua’s best offering, featuring 120mm fan support, a series of copper heat pipes and a promise from the makers that it will beat the best, it’s time to find out if their statement is true. Target price at TomsHW Preis-Engine is ~€49




Specifications :

Compatibility:

AMD: Athlon64/64+ Socket754/940/939

Intel Socket 478

Intel Socket LGA775

No Fan is included but supports 120mm (120x120x25mm & 120x120x35mm)
Supports installation of 2 fans (extra clips are included)

Heatsink + Fan Dimensions: 155x124x95mm
Weight: 700gram

In the Box :

Mounting hardware for each supported platform

A detailed manual

A tube of generic thermal goop

please note that the fan in the picture above is NOT included with this heatsink

For this review Noctua send along a new Coolink 120mm SWiF-1201 for testing, more info on these ex-“Noiseblocker” fans here.

Construction :

The NH-U 12 is what I would call a “tower heatsink” as it towers over the other components on your motherboard, even more so than its smaller brother; Fitted with 4 U-shaped heatpipes which are joined in a copper base, a series of aluminum fins dissipate the heat.





The base of the unit features holes for installation of different mounting brackets



The machine lapping traces are very hard to spot on the base, which is smooth but not to a mirror finish.



Installation next ->

Installation :

The NH-U 12 uses a custom mounting bracket, so you need to remove the stock AMD one, and thus remove your motherboard from the case. The Athlon 64 mounting brackets installed



It turns out that orientation of the NH-U 12 is not ideal when combined with this DFI motherboard, blowing hot air upwards. Unfortunately there is no way to rotate the heatsink towards the rear.



Fan mounting is through the use of 2 clips, installation is easy when you still have the motherboard outside the case. Once inside the case swapping fans can be a challenge, as this large heatsink will occupy the space needed to maneuver to get to the clips. Bridged or webbed fans are incompatible with these clips (example).

Noctua provides 4 strips of rubber to reduce fan vibration (2nd set is when you decide to install a 2nd 120mm fan on the heatsink).



And here is the NH-U 12 ready to go!




Performance and Noise :

Now the moment of truth, claiming to be the best I decided to compare the NH-U 12 results with those from Scythe’s powerhouse, the Shogun, in the graphs below. Before I start, please remember that the NH-U 12 is not orientated ideally, and you might see better performance when you can blow hot air towards the rear. The Shogun however can be installed either way, and in this test is orientated towards the rear.



And of to an excellent start, the NH-U 12 takes the 2nd place, only trailing the Shogun by a meager 2°C. The Shogun is slowly blowing hot air towards the rear… over the PWM area; this explains the higher temperatures.

The Coolink 120mm fan offers the exact same performance/ratio as my Papst super-silent 120mm fan!

Let’s remove more heat by adding extra case cooling:



More case airflow seems to help the NH-U 12, it now has a (very small) lead over the Shogun; I think it’s fair to say both heatsink deliver equal performance.



Again the same result, that is, there’s only a negligible difference between the two.


The Verdict:

Noctua came through on their claim; I compared their NH-U 12 to the best and it stood its ground. It’s excellent with very slow/quiet 120mm fans and that will interest those looking to silence their system and not lose performance. Recommended as and alternative to Scythe’s Shogun.

PRO
Compatible with all current AMD and Intel sockets
Top performance with both silent / powerful fans
Competitively priced
Possibility to install a 2nd fan (extra clips included!)

CON
Motherboard removal required
Can not change heatsink orientation

Where to buy?

Silverstone NT02:
Supplied by: Silverstone

Silverstone’s been in the system cooling for quite some time now, offering great design cases which feature excellent airflow. The NT02 is designed for high performance CPU Cooling, weighing in it at 880gr it’s no surprise the main ingredient is copper!

Target price at TomsHW Preis-Engine is ~€46




Specifications :

Compatibility:

AMD: Athlon64/64+ Socket754/940/939

Intel Socket 478

Intel Socket LGA775

Fan included: 80x80x25mm (a custom design with 92mm blades)
Fan Speed: 1800-3800rpm
Max Air Flow: ~74CFM

Heatsink Dimensions: 114x88x71mm
Weight: 880gram

In the Box :

Mounting hardware for each supported platform

A manual

A tube of generic thermal goop

A fan controller which allows you to half the fan’s speed

Construction :

The Silverstone Nitrogon 02, or NT02, is a small tower heatsink, three L formed heat pipes transfer heat from the copper base to the copper fins at the top, to increase cooling efficiency a shroud is placed over them to guide the air through. A smaller aluminum heatsink is placed on the base to help dissipate the heat.




The included fan is an Everflow fan which features 92mm blades fitted in an 80mm fan frame. Compact size, yet superior airflow to a standard 80mm fan:




Brackets are screwed to the base for installation




Although the base is quite smooth, I could clearly see machine lapping traces:




Installation next ->

Installation :

A custom mounting bracket supports this heavy heatsink, motherboard removal required. The fan is screwed on to the shroud, making swapping it almost impossible once it is installed. The fan gets power through a 4-pin molex, when connecting the fan controller you can reduce the speed quickly (because at default speeds it is very loud!)

The heatsink orientation was perfect with the DFI, blowing hot air backwards, unfortunately you can not rotate the NT02 in case your CPU socket is point the other way.




Performance and Noise :



Noise wise the NT02 is on par with the stock A64 HSF when running at 50%, when turning up the speed to 100% there’s a ~5°C temperature drop, not quite worth the 19dBA increase.



Adding more case cooling does a not change things a lot, I only saw a small temperature drop, and the difference between running at 50% vs 100% is still 5°C.



Again the NT02 offers the best noise/performance balance when its fan is running at 50%.

The Verdict:

The NT02 offers a noticeable improvement in performance over the stock A64 HSF but the included fan can not be qualified as a silent solution, even at 50%. Silverstone heatsink is capable of high performance but than you’ll be able to hear your PC in the next room for sure (59-61dBA!).

The NT02 has an attractive price point for an all copper compact heatsink with wide platform compatibility.

PRO
Compatible with all current AMD and Intel sockets
Respectable performance when used with slow(er) fans. Excellent performance with high noise fans.
Competitively priced for a full copper heatsink.
Includes a fan controller

CON
Motherboard removal required
Can not change heatsink orientation
Included fan extremely noisy at full speed

Where to buy?

Overview Specifications

	Fan Support	Fan Included?	Weight (gram)	Motherboard removal?	AMD	Intel
AC Freezer 64	Custom 92mm	Yes	460	No	S754/S939/S940	no
AC Freezer 64 Pro	Custom 100mm	Yes	528	No	S754/S939/S940	no
Noctua NH-U 12	2*120mm	No	700	Yes	S754/S939/S940	S478/S775
Noctua NH-U 9	80/92mm	No	570	Yes	S754/S939/S940	S478/S775
Scythe Samurai Z	92mm	Yes	360	No	S754/S939/S940	S478/S775
Scythe Shogun	120mm	Yes	790	Yes	S754/S939/S940	S478/S775
Sharkoon Red Shock	80/92/120mm	No	820	Yes	S754/S939/S940	S478/S775
Silverstone NT02	80mm	Yes	880	Yes	S754/S939/S940	S478/S775
Thermalright XP-120	120mm	No	370	No	S754/S939/S940	S478/S775
Zalman CNPS9500LED	Custom Size	Yes	530	No	S754/S939/S940	S478/S775

Reference heatsinks installed

Before I present you with a compilation of all the results in a series of (large) charts, here’s a quick rundown on the reference heatsinks used:

Arctic Cooling Freezer 64
Features easy clip on installation, is now being replaced by it’s Pro version and sold at +/- the same price (~€20). Unlike the Pro version, its orientation can not be changed.

Scythe Shogun
Top performing product from Scythe, comes with a 120mm which is not too good, but the heatsinks offers one of the most silent CPU cooling solutions when combined with the correct 120mm fan. Its price is higher than most, but you get what you pay for (~€60).

The heatsink can be orientated either way, but installation is not straight forward, and requires a bit of patience.

Zalman CNPS9500LED
- Provided by Bacata

Zalman’s latest heatsinks offers excellent performance, even at low fan speed; straight forward installation as it uses the default A64 bracket. The heatsink can be easily orientated either way. Priced at ~€50 it’s no really a budget friendly CPU cooling solution.

Thermalright XP-120
Thermalright’s XP-120 has been around for quite some time now, and can be found priced as low as €37!. XP-120 offers relatively easy installation with its integrated clip on system. It was one of the first (if not the first!) heatsinks to support 120mm fans and it reigned the silent CPU cooling market for quite some time. You’d be interested to know that they have announced their own ”tower” heatsink: HR-01. Let’s hope I can take it for a test run soon.





Let’s start our chart madness by sorting the results by Pure Performance ->

Pure Performance

When you don’t care about how noisy your system gets, and you just want to know which heatsink will keep your CPU running as cool as possible, these charts are for you !

---

For easiness of reference I’ll include the test system full specs, and details on each test setup. I won’t comment much on these results as the numbers speak for themselves. Do take into account inaccuracies of 1-2°C (and that’s thinking positive) in these results due to mounting differences and overall accuracy of my test material.

---

Results Test Setup 1

Antec Sonata II + Silverstone Passive PSU

AcoustiFan DustPROOF 120mm @ 5v in the rear as outtake (mounted with Acousti Fan Gasket)

Athlon 64 3200+ @ 2400 – 1.7v vcore

nVidia TNT2 Passive cooled video card

Noise produced with system running without HSF fan: 34dBA @ 50cm


In a low noise test setup, it’s no surprise that those heatsinks equipped with high speed fans take the lead.



Results Test Setup 2

Antec Sonata II + Antec SmartBlue PSU

AcoustiFan DustPROOF 120mm @ 7.5v in the rear as outtake (mounted with Acousti Fan Gasket)

Antec Sonata II’s 120mm fan in front @ 7v (mounted with Acousti Fan-Mounts)

Athlon 64 3200+ @ 2400 – 1.7v vcore

Chaintech FX5900XT with Arctic Cooling Silencer

Noise produced with system running without HSF fan: 50dBA @ 50cm


Only a 5°C difference in the TOP 19 entrees!


Results Test Setup 3

Antec Sonata II + Antec SmartBlue PSU

AcoustiFan DustPROOF 120mm @ 7.5v in the rear as outtake (mounted with Acousti Fan Gasket)

Antec Sonata II’s 120mm fan in front @ 7v (mounted with Acousti Fan-Mounts)

Athlon 64 3200+ @ 2200 – 1.5v vcore

Chaintech FX5900XT with Arctic Cooling Silencer

Noise produced with system running without HSF fan: 50dBA @ 50cm


Only a 5°C difference in the TOP 25 entrees!


Now let’s correlate CPU temperature to PWM temperature and noise production ->

Noise/Performance

Now how does each heatsink rank when I take the noise production into account? The PWM area should not run too hot either, so I’ve added this result, but since the PWM temperature is less important than the CPU temperature here’s the “formula” for this chart: CPUtemp+(PWMtemp/2)+Noise

---

For easiness of reference I’ll include the test system full specs, and details on each test setup. I won’t comment much on these results as the numbers speak for themselves. Do take into account inaccuracies of 1-2°C (and that’s thinking positive) in these results due to mounting differences and overall accuracy of my test material.

---

Results Test Setup 1

Antec Sonata II + Silverstone Passive PSU

AcoustiFan DustPROOF 120mm @ 5v in the rear as outtake (mounted with Acousti Fan Gasket)

Athlon 64 3200+ @ 2400 – 1.7v vcore

nVidia TNT2 Passive cooled video card

Noise produced with system running without HSF fan: 34dBA @ 50cm





Results Test Setup 2

Antec Sonata II + Antec SmartBlue PSU

AcoustiFan DustPROOF 120mm @ 7.5v in the rear as outtake (mounted with Acousti Fan Gasket)

Antec Sonata II’s 120mm fan in front @ 7v (mounted with Acousti Fan-Mounts)

Athlon 64 3200+ @ 2400 – 1.7v vcore

Chaintech FX5900XT with Arctic Cooling Silencer

Noise produced with system running without HSF fan: 50dBA @ 50cm




Results Test Setup 3

Antec Sonata II + Antec SmartBlue PSU

AcoustiFan DustPROOF 120mm @ 7.5v in the rear as outtake (mounted with Acousti Fan Gasket)

Antec Sonata II’s 120mm fan in front @ 7v (mounted with Acousti Fan-Mounts)

Athlon 64 3200+ @ 2200 – 1.5v vcore

Chaintech FX5900XT with Arctic Cooling Silencer

Noise produced with system running without HSF fan: 50dBA @ 50cm




Now I know what you are thinking, those results in test setup 2 and 3 are mostly rendered useless because of the overly noisy Antec PSU, at 50dBA a lot of heatsinks fall well below this threshold. But don’t worry I have the solution ->

Noise/Performance : Alternative chart for Test 2 & 3

How to overcome the noise produced by the Antec PSU, well since I did record the noise levels of each fan in the first test setup which has a “noise floor” of ~34dBA, I’ll use those values in my charts for test setup 2 and 3. (Had to retest 92mm Vantec and 120mm Titan in setup 1, but as you’ll see, they are still very loud.. only a ~2dBA drop in noise)

By making this adjustment I negate the noise added by the Antec PSU and extra case fans, putting the focus on the heatsink’s noise production, in this heatsink roundup.

The “formula” for this chart remains the same: CPUtemp+(PWMtemp/2)+Noise

---

I won’t comment much on these results as the numbers speak for themselves. Do take into account inaccuracies of 1-2°C (and that’s thinking positive) in these results due to mounting differences and overall accuracy of my test material.

---

Results Test Setup 2




Results Test Setup 3






Conclusion

What did I learn after more than 140 hours of testing and data processing? Testing inside a case reduces the performance differences between the heatsinks a lot!. Almost all the high end heatsink tested today deliver comparable performance, which is of course a good thing for you, the reader, as it will be hard to make a “wrong purchase”.

I was also able to experiment to see if adding more/less case cooling impact CPU cooling, and although it does so, it was mostly the PWM area which benefited the most from extra airflow.

It’s impossible to come up with a global “roundup” winner, as each of these heatsinks has their benefits and drawbacks. To see these check the “Install/Test” pages (easily accessible through the Quick Jump menu at the top&bottom).

I'm looking forward to testing new heatsinks as they arrive here at "the lab" (if you can call it that). The Thermaltake Big Typhoon seems to be a popular choice among enthusiasts, as well as the new Thermalright HR-01.

Question/Comments/Remarks? Please share them at our forums