New Elite Products vs. Intel Core 2 Extreme

Today we are to examine more good-looking and expensive coolers with heat pipes from famous manufacturers. These coolers will again fight for the right to cool the elite model of the top desktop dual-core processor from Intel based on the victorious Conroe core.

Summary of the previous article. Our previous articles

• "Core 2 Extreme vs. Zalman CNPS9700 LED and CNPS9500 AT and Others. or an Extreme Processor Needs an Extreme Cooler" and
• "Cooler Master Mars, Hyper UC, Hyper TX and Others vs. Intel Core 2 Extreme — Aluminum vs. Copper and Heat Pipes Again"

reviewed some advanced coolers with heat pipes from the elite manufacturers of coolers. But our task was not only to try the coolers proper - we liked very much Zalman CNPS9700 LED and Cooler Master Hyper TX. We were also to find out what reefs awaited users, who bought new Core 2 Duo/Extreme processors from Intel and decided not to pay much attention to cooling, setting hopes upon their "40% as low" power consumption. As we have already noted in previous articles, the stumbling block of some coolers in case of Core 2 Duo and Core 2 Extreme processors is the maximum admissible (according to Intel specs) temperature inside the heat spreader of these processors - just 60.1°C-61.4°C. It's nearly 10°C lower than in the previous generations of Intel processors - Pentium 4, Pentium D and the like. That is this 40% power saving of Intel Core 2 versus Pentium D, advertised by the corporation, is practically "gobbled" by harder requirements to the maximum admissible temperature. New Intel processors need no less efficient coolers than their much more voracious predecessors - Pentium 4 and Pentium D.

First of all, we are going to examine a top model from Cooler Master - Eclipse, which didn't make it to our previous shootout of products with heat pipes from this manufacturer. At multiple requests of some of constant guests of our restaurant "Food for Thought" :), the second "course" in our today's exquisite menu is the good old Thermaltake Big Typhoon in a new incarnation - Big Typhoon VX.

Besides, we are going to clear up the situation with "right" and "wrong" orientation of coolers with heat pipes, because some "gourmets" have "wrong" opinions about them. This section is added at the request of some constant readers. :)

Cooler Master Eclipse (RR-CCB-WLU1-GP)

As we have already said, this cooler is currently a top model in this series (that is a main product "offered by a chef" on the manufacturer's web site) of all-purpose (K8/LGA775) air coolers from this company.

I don't know why this product was given such a questionable name (eclipse) - reviewers were given an opportunity to chitchat ;)). But I have an impression that Cooler Master engineers also went into a similar phase, designing products (like Mars), which originality and "coolness" of the exterior predominates over the care for efficiency and users' purse. :) However, the manufacturer published the following extract, as if to justify the name and design of the Eclipse cooler:

"The look is different, because this cooler is different. Cooler Master's brand new cooler – Eclipse is the realization of a lifestyle that integrates style and function, beauty and technology. Eclipse's exceptional design of a gliding fan duct impeccably channels the airflow to cool down not only CPU itself, but also the surrounding components (VRM, Memory, NB Chipset, and etc)."

I cannot tell you how new and different the idea with a turbine is, but I remember similar (and well forgotten ;)) products.

However, Eclipse certainly has a point, as its 4.5 mm copper base is connected to the heatsink (consisting of 0.5 mm fins) not only with four standard 6 mm heat pipes, but also directly - that is some fins touch the base of the cooler. Unfortunately, the heat pipes are just pressed into the copper of the base and fins, plus 1 mm edging of fin holes. You can easily move the fins with a finger in contact places (I found no traces of glue). This method is easy, of course. It's used by many manufacturers (including some expensive coolers from this manufacturer), but the thermal contact is not the best.

Alas, the surface area of Eclipse finning is not included into its specs. To calculate it on your own is a good task for entry math exams due to the complex configuration of most fins. :) Heatsink dimensions — 132x120x105 mm. The cooler weighs much (670 g), even though it uses aluminum. The "eclipse" nature of the design is highlighted by a transparent dark "semi-housing" that directs some of the upward exhaust airflow down (God knows where it actually directs the air). You can change the slope of this housing. Cooler dimensions with the housing — 147x146x110 mm.

The non-standard turbine blower (66x68 mm, long-life bearing, 40,000 hours of service life) is equipped with cylindrically bent blades. It's designed for up to 580 mA and has a 4-pin connector (for PWM control) with an additional 3-pin speed jumper. Its maximum and minimum rotational speed (according to specifications) in manual mode (jumper) is 1800 rpm and 3300 rpm correspondingly. But in case of PWM control, rotational speed can change from 900 rpm. Performance of the fan at maximum speed is 39.8 CFM. The noise at minimum speed is 17 dBA. Unfortunately, this design of a speed control is inconvenient to install on the front or rear panels, you'll have to use extension wires and switches.

The cooler is mounted on the motherboard with a standard clip (for AMD processors, see the photo above) that goes through two slits in heatsink fins, or with a frame secured with four screws (for all LGA775 processors from Intel). In the latter case, the counter frame that does not let a motherboard to bend is not included into the bundle. In return, it includes a six-sided adapter to screw nuts with a usual screwdriver. The manufacturer guarantees interference-free installation. But in practice, we could install Eclipse on the Intel D975XBX motherboard only in one orientation, because three other positions were impossible due to interferences from the heatsink, retention frame, and PCB elements.

The cooler comes shipped in a large box (the largest I've ever seen) with lots of inscriptions (specs, a list of supported processors from AMD, a Intel, etc). The bundle includes everything necessary, including a tube with brand thermal compound (the base is protected with a sticker). It's a RoHS compatible cooler, that is it's manufactured lead free.

Thermaltake Big Typhoon VX (CL-P0310)

The distinguished Big Typhoon does not need our recommendations — since this design first appeared on the market over a year ago (see our review), lots of "gourmets" liked it and installed it in their system units.

Despite its huge dimensions, this all-purpose model (K8/LGA775) has a reasonable price ($20 cheaper than Eclipse) and offers excellent characteristics that put it in line with top competing products.

The 122x122x103 mm heatsink consists of a built-up copper base (50x50x9 mm) with six 6 mm copper heat pipes (3 pipes at each side) and 142 aluminum 0.3 mm fins at 1 mm steps (such gaps can get cluttered with dust, and the fins are not arranged evenly) with the total surface area of about 7000 sq. cm. The pipes are pressed and glued into the fins. The cooler weighs 827 g.

A big 12 cm fan TT-1225A of the proprietary Thermaltake color blows not only at the fins and the base, but also at motherboard components around the socket. The fan consumes no more than 250 mA. A miniature variable resistor installed in one of the corners of the air grille (the minimum output voltage is 7 V) can control the rotational speed from 1300 rpm to 2000 rpm. Though relatively small, this fan provides up to 86.5 CFM at 16-24 dBA noise.

The updated VX model differs from the old Big Typhoon

• not only in a faster (at maximum speed) fan and
• a built-in speed control, but also in
• a retention module.

The fact is, the retention module of the old Big Typhoon gave rise to unfavorable criticism — too difficult to install, the cooler could be installed askew, which would damage thermal contact. The present solution is simpler: there is a metal frame for LGA775 with standard rotating pistons (like in a boxed cooler); AMD K8 processors will use a pressure lever (unfortunately it uses only one latch at each side, which may lead to chippings or weak contact). Nevertheless, the new retention module made the installation procedure much simpler. By the way, design of the cooler is so good (compact at the base) that it allows to install it on most motherboards in any of the four orientations.

Big Typhoon VX comes shipped in a red-and-black cardboard box with a transparent insert and a list of characteristics, specifications and key advantages. The bundle includes two assembled retention modules (clips), white liquid thermal compound in a tube and a brief installation manual. We'll use this cooler to test heat release in three typical orientations of heat pipes.

Procedure for Evaluating Heat Transfer

As we did before, we used Intel Core 2 Extreme X6800 (2.93 GHz) operating at 3.20 GHz (we raised CPU multiplier by one step). FSB and memory frequency on the Intel D975XBX motherboard were set to nominal 266 (533/1067) MHz. The processor dissipates <82W under these conditions, like a hypothetic dual-core CPU, the next in line after the X6800. Besides, 3.2 GHz is a guaranteed level, to which Conroe processors can be overclocked without any problems. So we also evaluate "capacity" of these coolers for overclocking Core 2 processors. Our contenders also include all coolers with heat pipes from two of our previous reviews (see the links in the beginning of the article).

The motherboard with the processor and a gigabyte of DDR2 memory were installed inside a closed PC case (midi-ATX Palo Alto PA-810) with the HiPro HP-W460GC31 PSU (460 W), ASUS AX800 XT (ATI X800 XT) video card, WD800JD HDD, and one 90 mm fan to drive the air inside at 2500 rpm. We used liquid Zalman ZM-STG1 thermal compound for all coolers, except Big Typhoon VX (it was tested with the native thermal compound). Room temperature during our tests was kept at 22°C. Temperatures of the processor and motherboard (near the CPU voltage regulator) were read by built-in thermal sensors and Everest Ultimate Edition 3.01.652, SpeedFan 4.30 and Intel DCC under MS Windows XP SP2. We also registered the rotational speed of the CPU cooler. CPU was loaded by the Conroe-optimized S&M 1.8.1 on three different levels:

• 100% (maximum temperature, extremely rare in real life),
• 75% (according to the author of S&M, it's a typical gaming level), and
• 50% (according to the author of S&M, it's a typical office load, though this statement should be tested).

Of course, we also took readings of the idle system (0% load). EIST was enabled during the tests, because it corresponds to reality (we are trying to simulate reality here, not just bare temperature readings). That's why the processor dropped its clock rate to 1.6 GHz when its was idle. This situation is illustrated by RMClock v 2.15. It was used to make sure there was no throttling in the course of our tests (we disabled the standard tools of activating throttling for the time of our tests).

Cooler Master Eclipse was tested in three fan modes: at +12V, at constant reduced speed (jumper=2-3), and at the lowest speed (open jumper). Thermaltake Big Typhoon VX was tested with the speed control set to minimum and maximum.

Does an orientation matter?

So, first of all we shall see whether heat release of a cooler with heat pipes depends on its orientation in space, relative to the gravity vector. Many people still think that it does, because steam is lighter than liquid. ;)

Big Typhoon VX is rather convenient for this test. It can be installed in various positions. Besides, it's equipped with a speed control. Big Typhoon VX was tested in three orientations in space:

• Motherboard is installed vertically, heat pipes are in a vertical plane ("vertic")
• Motherboard is installed vertically, the heat pipes are in a horizontal plane ("horiz")
• The motherboard is installed horizontally (the PC case lies sidelong, no label)

We hope to monitor how heat release of a cooler depends on the orientation of heat pipes in gravity. ;) Let's have a look at diagrams with test results for two rotational speeds.

There is evidently no dependence on the orientation, the cooler works well (within the measurement error) in all modes. Why? That's because heat pipe manufacturers understand the need to get rid of gravitational "anisotropy". For this purpose, they put porous material inside heat pipes. So the liquid inside moves equally well along and against gravity due to the capillary effect, evaporating at the hot end of the pipe, wherever it is. :)

By the way, these diagrams show well that a reduction in rotational speed of Big Typhoon VX by one third (resulting in lower noise, see below) almost does not deteriorate its heat release — the temperature of cooled components rises by just one degree! That is the cooler is close to its maximum efficiency even at 1300 rpm (the bottleneck of this design at a higher rotational speed is heat pipes, their contact with the base and fins, as well as insufficient airflow between fins — perhaps they should be placed at larger steps). So there is no point in increasing the rotational speed of this cooler to 2000 rpm. :)

Results of Heat Release Tests

Now we proceed to the cooler shootout. Let's start with maximum load (100%) of the dual-core processor in S&M. It's extremely rare in practice for considerable amounts of time.

Even in such heavy conditions, practically all contenders with heat pipes cope with cooling the overclocked processor to the temperature published in its specifications (temperature measured inside the CPU die is higher by several degrees than the temperature at the heat spreader, specified by Intel). Mars from Cooler Master actually performs a tad worse than the others. But the latest Eclipse put up a good show like much cheaper CM Hyper TX and GT Igloo 5700 MC. Big Typhoon is a joy. It demonstrated results on the level of more expensive Zalman CNPS9500 AT. It was outperformed only by the super elite Zalman 9700! Moreover, Big Typhoon VX is the best model, when the rotational speed is reduced to 1300 rpm. Its thermal efficiency was practically no worse (see above), it significantly outperformed even Zalman CNPS9700 LED at the same fan speed! By the way, Big Typhoon VX is one of the best coolers for PCB elements near a processor.

If we switch to a much more realistic CPU load (75% load according to S&M, see the diagram above), the cooler ranks remain practically the same: Big Typhoon VX is still the third at full fan speed. It competes with the CNPS9500 AT. It's the best cooler at reduced speed, outperforming much the CNPS9700 LED - it provides best cooling for PCB elements. Cooler Master Eclipse performs well, being on a par with the CM Hyper TX and GT Igloo 5700 MC at full speed. It's one of few coolers that can cool a processor to the standard temperatures at a reduced fan speed.

The situation gets much easier, when this powerful processor is used under unusual "half-load" (50% by S&M, Diagram 3). Practically any of our contenders at reduced fan speed can cope with the Conroe processor overclocked to 3.2 GHz. The same leaders here.

When idle (Diagram 4), we shouldn't worry about the core temperature of a Core 2 processor at all, if your system is equipped with an expensive cooler. :):):)

Acoustic tests

We measured the noise generated by coolers using the VSV-003-M3 sound level meter according to our test procedure (see here and here). Results are published on the diagram together with fan's tachometer readings.

The quietest cooler at reduced fan speed (PWM mode) is Cooler Master Eclipse with a small turbine. It outperforms even Zalman CNPS9500 AT operating at 660 rpm! Unfortunately, operating in this noiseless mode, it can cope only with the office load (50%) of our processor. While in case of the constant medium speed (1900 rpm) Eclipse actually copes (in our conditions - ventilated PC case) even with 100% S&M load of the Conroe 3.2 GHz, generating 26 dBA noise (relatively quiet, but still audible). Big Typhoon VX at minimal speed (1300 rpm) generates as much noise. But its thermal efficiency is much higher! Considering that the noise from Big Typhoon at full speed grows to 38 dBA, while its thermal efficiency does not change much, I think there is no point in using this operating mode. :) The "extra charge" for a more powerful fan and a speed control is not expedient. By the way, Cooler Master Eclipse at full speed is the noisiest contender here. Considering the heat release data published above, there is also no point in using it in this mode, you should use it at the fixed medium speed or with PWM control.

Conclusion

So, we've examined another new product from Cooler Master — Eclipse with four heat pipes and a turbine, and Big Typhoon VX from Thermaltake (the latest modification of a one-year-old cooler). Unfortunately, Eclipse is similar to Mars (it demonstrates moderate heat release, but offers cool exterior and comes at a high price). It didn't follow the praiseworthy tradition of the semi-budget Hyper TX series. In particular, Eclipse is no better than Hyper TX in thermal efficiency, though it's much noisier (in similar modes). Can we recommend the $60 Eclipse, if a user can buy the $27 Hyper TX from the same manufacturer or even a more efficient $40 Big Typhoon VX from Thermaltake (we don't mention Zalman coolers here)? The answer is obvious, I think. Big Typhoon VX is another pair of shoes. It confirmed its top efficiency, though its design is not devoid of drawbacks: it's just inexpedient to use it at fan speed exceeding 1300 rpm.

Write a comment below. No registration needed!