iXBT Labs - Computer Hardware in Detail

Platform

Video

Multimedia

Mobile

Other

Five Premium CPU Coolers In Tests

From ASUS, Noctua, Scythe, Thermalright and Thermaltake.

June 26, 2009



<< Previous page

     Next page >>

Several questions about the usage of these coolers arose during our tests, and this part of the article will try to answer them.

Question 1: If noise ergonomics is the first priority, then what's the effectiveness and practicability of two-fan assemblies of Scythe Mugen 2 and Thermalright IFX-14?

Diagram 11. Temperature readings (CPU core temperature, two-fan assemblies of Scythe Mugen 2)


As we can see, comparison of 'equally-noisy' combinations of stock and two-fan modifications of Scythe Mugen 2 (with Scythe Slip Stream fans) is somewhat ambiguous. The only interesting assembly here is the one operating at 900rpm (blowing fan) and 500rpm (exhaust fan) -- at the same noise level, it allows to get a small thermal 'discount' versus a single 900rpm fan, which cannot provide enough air pressure to blow through the optimized, but still large fin-stack of Mugen 2. The other modifications are either of little use (1300+900rpm), or useless (1900+1300rpm), or even harmful (1900+1900rpm). By the way, the most powerful assembly (1900+1900rpm) illustrates thermal limits of the Mugen 2 -- a single 1900rpm fan is more than enough for it, higher airflow pressures yield no effect but nasty noises generated by the high-speed exhaust fan.

Answer A: Scythe Mugen 2 copes with its responsibilities brilliantly even with a single fan. Results of two-fan assemblies are of little use in practice, they are not expedient.

Diagram 12. Temperature readings (CPU core temperature, multi-fan assemblies of Thermalright IFX-14)


The situation is different here -- the second fan mounted inside the Thermalright IFX-14 heatsink between fin-stacks is very helpful. The most interesting combination in our opinion is with slow Scythe Slip Stream fans (500+500rpm) -- noise grows insignificantly (you may even fail to hear any changes in background noise, because registered levels of acoustic pressure are very low), but temperatures drop considerably. It's only natural -- an additional fan actually 'activates' the second fin-stack of the IFX-14 (with a single low-speed fan mounted on the outer side of the first fin-stack, the second fin-stack actually works in conditions of natural convection, that is in passive mode). And overclockers may install two high-speed fans. For example, an assembly with two Minebea 1900rpm fans achieves very low thermal resistance for air coolers (0.16 °C/W!), which is similar to results of liquid cooling devices (with powerful pumps and two-fan heatsinks).

Answer B: The efficiency of two-fan assemblies of Thermalright IFX-14 is very high.

On the other hand, three fans in IFX-14 assemblies look excessive and even harmful, at least if we speak of extreme options -- slow and fast models. Don't bite at marketing tricks.

Question 2: What is the most optimal way to install a single fan in Thermalright IFX-14 -- on the outer side or in between the sections?

Diagram 13. Temperature readings (CPU core temperature, fan install options in Thermalright IFX-14)


It's an illustrative situation with a clear conclusion -- a slow fan must be installed on the outer side. There is nothing surprising here. If a slow fan is installed in between fin-stacks, its efficiency gets much worse: there appear problems with ventilation of the front section (insufficient air pressure, because air is taken not through the fin-stack, but from fan periphery -- air passes around the front section, leaving it passive), hampered ventilation of the back section (air temperature in the center between the fin-stacks grows closer to 50°C -- that is, even mixed with cooler peripheral air at the input, the back section is always at a thermal disadvantage). As a result, temperature grows (8°C with a Slip Stream 500rpm and 5°C with a Minebea 500rpm). Nevertheless, a powerful fan achieves a much more productive balance in exchanging heat on the fin-stacks -- with the Minebea 1900rpm installed 'inside', Thermalright IFX-14 demonstrates even higher efficiency than with an external fan.

Answer: In low-noise modes, the best solution is to install a fan on the outer side of the Thermalright IFX-14. But if you want to use a high-speed and high-pressure fan (over 1900rpm), you should install it between the fin-stacks.

Question 3: What advantages does the bundled HR-10 unit provide for Thermalright IFX-14?

Diagram 14. Temperature readings (CPU core temperature, Thermalright IFX-14 and HR-10)


Well, test results speak for themselves -- thermal efficiency of the IFX-14 is not affected by this unit at all. It's a predictable outcome: collateral heat release along the line 'processor-motherboard' for the Intel LGA775 platform is very low, the same insignificant effect on the socket is demonstrated by near-socket motherboard components. So even if there is some effect from external factors, it's absolutely insignificant for a CPU. Results of our testbed do not automatically apply to all scenarios. But we risk an assumption that the situation will be similar in most users' systems.

Answer A: Alas, you cannot count on lower CPU temperatures, when you use the IFX-14 together with the HR-10 unit.

Nevertheless, there is some positive effect from installing the HR-10 unit, and it has to do with temperature of near-socket components of the CPU voltage regulator.

Diagram 15. Temperature readings (temperature of chokes, Thermalright IFX-14 and HR-10)


As we can see, HR-10 performs well in noiseless assemblies of the IFX-14, when ventilation of the near-socket components is insufficient. It makes quite a difference in temperatures of the CPU voltage regulator here -- for example, the temperature difference on chokes PL26 and PL25 for the assembly with a 500rpm Slip Stream reaches 7-8°C, which corresponds to the fan of a higher rank (much noisier Minebea 700rpm). However, as ventilation of the socket improves, the effect of this unit grows noticeably weaker -- in case of the reference fan Minebea 700rpm, results of assemblies with and without HR-10 become very close to each other, and they are absolutely identical for the ergonomic mode (1300rpm).

Answer B: The HR-10 unit contributes to normalization of temperature conditions of near-socket components on the motherboard, but positive effect from its usage is limited only to low-noise assemblies of the IFX-14.

Question 4: If noise ergonomics is the first priority, then what's the effect of additional ventilation on results of our coolers? Let's answer this on the example of Scythe Mugen 2 and Thermalright IFX-14.

In order to answer this question, we simulated a typical situation in popular PC enclosures with a PSU installed on the bottom and two fans (rear and top), which are usually mounted on the rear and top walls right opposite to the cooler. We installed two 300rpm Scythe Slip Stream fans -- one of them as a rear exhaust fan (mounted flush with the outer surface of IO-connectors on the motherboard), the other was installed 6cm from the upper edge of the motherboard to pump air in. We tested two options -- with only the rear fan working and with two fans.

Diagram 16. Temperature readings (CPU core temperature, Scythe Mugen 2 assemblies)


The most effective ways to provide extra ventilation are expectedly noiseless assemblies of Mugen 2 (500rpm fan) -- adding a rear fan drops the temperature by 4°C, the second fan on top drops another 1°C, without changing the background noise. On the other hand, it's useless to install slow fans for mid and high-speed configurations of a cooler (including the nominal 1300rpm assembly).

Answer A: Noiseless assemblies of Scythe Mugen 2 benefit much from installing a rear exhaust fan.

Diagram 17. Temperature readings (CPU core temperature, Thermalright IFX-14 assemblies)


We got more interesting results here -- the effect of the rear fan is low (even with the weakest assemblies of the IFX-14), but installation of the second blower fan on top improves temperature conditions. That's understandable. A rear fan alone fails to cope with ventilation of the corresponding fin-stack (it's an obstacle rather than a useful load). The front part of the fin-stack with a quiet fan still works independently (the outer side fan barely blows through its own segment of the fin-stack, it just fails to reach the back segment). But when the second fan is installed on top, everything falls into place -- the cooler gets a gulp of fresh air, which creates necessary conditions for effective heat exchange in both fin-stacks. However, as in case of Scythe Mugen 2, ergonomic and high-speed assemblies of the IFX-14 are indifferent to low-noise fans.

Answer B: Noiseless assemblies of Thermalright IFX-14 benefit much from installing a top blower fan.

As for the other coolers, we don't think it's expedient to examine how they work with additional low-speed fans, because their results in these test modes remain the same (except for the noiseless assembly of ASUS Royal Knight, which gets a measly 1°C discount from the top fan).

Conclusions

ASUS Royal Knight. A decent product with interesting functionality. It shows hi-end results practically in all modes (especially low-noise), and it can boast of good technical quality as well as friendly usability. The only drawback of this cooler is its price (about $70), it's too high in our opinion.

Noctua NH-C12P. This product product follows the successful NH-U12P -- it demonstrates the same unprecedentedly high technical quality, spiced up with competitive performance and neat usability. Probably the most promising computers for the NH-C12P are hi-end HTPCs, where it can reveal its full potential.

Scythe Mugen 2. A capable new product demonstrating champion results. This cooler is a direct competitor of Thermalright IFX-14, showing advantage in major technical aspects (dimensions and weight in the first place). If retail prices for Mugen 2 are just as sweet as promised by the manufacturer (about $50), this cooler may become an absolute hit.

Thermalright IFX-14. The combination of its technical and functional features makes it the best modern cooler. It may be a perfect choice for both overclockers and users who prefer quiet computers. But the price tag may scare users away -- the bundle with two excellent fans costs $120-140 in our locality.

Thermaltake BigTyp 14Pro. An ambiguous product. On the one hand, BigTyp 14Pro cannot offer decent efficiency in noiseless modes (where this cooler is only a tad better than the old Big Typhoon). But on the other hand, in ergonomic modes it demonstrates a good combination of quality and functionality. Quite a dilemma. And while we don't doubt that BigTyp 14Pro is a true hi-end cooler, the time will tell if it can win the hearts of computer enthusiasts.


ASUS Royal Knight provided by ASUS,
Noctua NH-C12P provided by Noctua,
Scythe Mugen 2 provided by Scythe.

Write a comment below. No registration needed!


<< Previous page



blog comments powered by Disqus

  Most Popular Reviews More    RSS  

AMD Phenom II X4 955, Phenom II X4 960T, Phenom II X6 1075T, and Intel Pentium G2120, Core i3-3220, Core i5-3330 Processors

Comparing old, cheap solutions from AMD with new, budget offerings from Intel.
February 1, 2013 · Processor Roundups

Inno3D GeForce GTX 670 iChill, Inno3D GeForce GTX 660 Ti Graphics Cards

A couple of mid-range adapters with original cooling systems.
January 30, 2013 · Video cards: NVIDIA GPUs

Creative Sound Blaster X-Fi Surround 5.1

An external X-Fi solution in tests.
September 9, 2008 · Sound Cards

AMD FX-8350 Processor

The first worthwhile Piledriver CPU.
September 11, 2012 · Processors: AMD

Consumed Power, Energy Consumption: Ivy Bridge vs. Sandy Bridge

Trying out the new method.
September 18, 2012 · Processors: Intel
  Latest Reviews More    RSS  

i3DSpeed, September 2013

Retested all graphics cards with the new drivers.
Oct 18, 2013 · 3Digests

i3DSpeed, August 2013

Added new benchmarks: BioShock Infinite and Metro: Last Light.
Sep 06, 2013 · 3Digests

i3DSpeed, July 2013

Added the test results of NVIDIA GeForce GTX 760 and AMD Radeon HD 7730.
Aug 05, 2013 · 3Digests

Gainward GeForce GTX 650 Ti BOOST 2GB Golden Sample Graphics Card

An excellent hybrid of GeForce GTX 650 Ti and GeForce GTX 660.
Jun 24, 2013 · Video cards: NVIDIA GPUs

i3DSpeed, May 2013

Added the test results of NVIDIA GeForce GTX 770/780.
Jun 03, 2013 · 3Digests
  Latest News More    RSS  

Platform  ·  Video  ·  Multimedia  ·  Mobile  ·  Other  ||  About us & Privacy policy  ·  Twitter  ·  Facebook


Copyright © Byrds Research & Publishing, Ltd., 1997–2011. All rights reserved.