Sign in with
Sign up | Sign in
Your question

Cool and Quiet HSF Review: Scythe/Zalman/Noctua

Last response: in Overclocking
Share
April 15, 2007 1:45:13 AM

Cool and Quiet. A tall order for a computer system, to be sure. Quite a few visitors to these forums come looking for help, trying to decide which CPU heatsinks to consider purchasing. Some look for the best performance, some want quiet, and some want a good balance of both. This is what this review and comparison is specifically meant to address: performance cooling that is also quiet. Often performance comes with a price, one of which is Noise. They are usually inversely proportional: the better the performance, the louder a heatsink generally is. There are exceptions: loud and hot (who wants that?) and cool and quiet. We’re hunting for the latter………

I am a bit of a performance junkie, but am also a noise nut too. In a perfect world, I want amazing performance that approaches silence, and at an affordable price to boot. Who doesn’t? But achieving it is difficult, compromises need to be made, extra money often spent, patience required, and so on. So I decided to take a look at a few popular CPU heatsinks to examine their real-world performance for Cool and Quiet.

I have limited resources (both time and money) and am currently unable to do as an exhaustive test list as I’d like. However, this review is an attempt to tackle a few products, popular ones, and hopefully help some people who struggle with achieving both performance and quiet for their computers. So, let’s review the contenders:

-Zalman 9500
-Scythe Ninja
-Noctua NH-U12F

3 popular products known for being quiet. But how about performance, not to mention mounting, quality of design, and price? Well, let’s find out…….

Before I continue, I’d like to thank Big Alcattle, a forum regular here, for graciously providing me with the Noctua and Scythe for testing purposes.


The purpose of this review is not to inundate people with countless specifications, photos, product information, etc. That is easily found on their respective company websites and product literature. Rather, the purpose here is to give a concise synopsis of the overall performance of the products under real world conditions. I wanted to approximate what would be a realistic setup for many people at home. So, no open-air test bench for this review ;) 

I’m primarily examining 3 main components to each product:

-Product Design
-Noise
-Temperatures

Let’s begin, shall we?


SETUP
With a real-world scenario in mind, the test setup was designed to approximate a fairly common setup for a higher-end computer system that strives for performance and noise reduction. Here’s the system specs: A Core2Duo CPU (stock and highly overclocked), and 8800 graphics card, and all your other standard fare. Below is the full specs:



The heatsinks were installed in the P180 case and the side panels were put on. Now, to make this a little more interesting, I placed my case close to the furnace vent (radiant heat). This accomplishes two things: first, it approximates winter conditions that many people experience in northern climates with the furnace on and ambient temperatures higher (but also approximates conditions for people that live in hotter southern climates as well); second, it amplifies the amount of heat and exchange the heatsinks have to dissipate in a closed case that experiences higher ambient temperatures…..meaning that the heatsinks will have to perform VERY well to maintain acceptable CPU temperatures.

Ambient temperatures were monitored with a digital thermometer, accurate to within 1 degree Celcius. Constant ambient temperatures were monitored during the tests to ensure consistent test setups for all products. Idle temperatures were monitored 20 minutes after computer startup, with no applications loaded other than normal Windows system processes. Load temperatures were monitored after 20 minutes of dual instances of Prime95. Highest recorded temperatures were noted for the tests.


PRODUCT DESIGN
Aesthetics
The Noctua and Scythe are considered “tower-style” heatsinks; that is, they rise above the motherboard like a tower. The Zalman’s design is a more of a “tunnel” that rises off the motherboard. All products are quite weighty, large, and imposing. Aesthetics are an individual taste, but many people seem to appreciate the look of the Zalman…..it has an LED light behind the 92mm fan, with a nice contrasting metal color that is generally well received, especially for cases with a side panel window. The Noctua is a more “crisp” tower design, very sleek and minimalist, and the 120mm Noctua fan offers a unique color combination for a case. The Scythe Ninja is the most bulky of the three, with an interesting cutout design that looks more “beefy” rising off the motherboard.

Mounting
The mounting design for the Zalman is fairly straightforward: a backplate, a pressure bar, and 2 screws that allow the heatsink to be installed facing the rear or top of the case in a Socket 775 motherboard. In fact, all 3 products can be positioned to face the rear or top, a very handy feature. The Zalman pressure bar is a bit awkward to get into proper position, but after a short bit of finicking, it sets in place. Getting the screws installed is straightforward. Overall, not very difficult, allows some minor twisting of the heatsink itself, fastens tight, and quite sturdy.

The mounting design for the Scythe Ninja is very simple: pushpin fasteners that twist and lock into place when installed. No backplate is required, and this can be a good thing if your motherboard is already installed in your case. The Ninja is more bulky at the base than the Zalman, and can be a tight squeeze into many mid-tower cases. It just clears the RAM sticks when installed, but overall there is adequate room available. One of the mounting pushpins can be difficult to turn and lock into place if you have a northbridge cooler that sits close to the edge of the heatsink…..not a big deal, but worthy of note. There is, however, one major design flaw of the Ninja’s mounting design: no backplate. The pushpin mounts are made of plastic, not metal, and this can pose some significant risks. The first is that a beginner may not know if they are fully installed, thus leaving the inadequate pressure between the heatsink and the CPU itself, thereby reducing thermal transfer and creating high temperature problems. The second concern is risk of breaking the plastic components. Considering the pressure required to install and hold the heatsink in place, this is a distinct possibility. Lastly, given the weight of this heatsink, I would be extremely hesistant to subject a case containing a Ninja to any sort of unsupervised or frequent transportation. Quite simply, there is too much risk of a mounting failure. In my opinion, this mounting design is a critical flaw in the Ninja’s overall design and performance.

The mounting design for the Noctua is also fairly straightforward: a backplate, 2 mounting plates and bars, and 2 spring-loaded mounting screws to hold it in place. Whereas the Zalman’s pressure bar was rather tricky to slot through the heatpipes, the Noctua is much easier.....there are, infact, 2 separate mounting plates that simply screw to the heatsink itself. The 2 mounting bars are screwed directly into the backplate and then the heatsink is screwed to the mounting bars with spring-loaded screws that stop when fully tightened…no guesswork on how hard to turn. Very clean, very simple, little effort required. The mounting design of the Noctua was the easiest of the three, and seemed the most stable and sturdy. I would have no hesitation transporting a case with the Nocua over long distances (even via courier shipment if necessary).


NOISE
I do not, unfortunately, have access to the sensitive (and expensive) measuring equipment that Tech website reviewers are able to use. However, since this is a real-world review, I will simply comment on my overall impressions of each of the heatsink fans, as compared to each other.

Both the Scythe and Noctua employ 120mm fans, whereas the Zalman uses a 92mm fan. That clearly puts the Zalman at a disadvantage, even before the computer is turned on. However, the Zalman is the only heatsink of the three that comes with a fan controller, to find a noise level that is acceptable for each person.

The Scythe Ninja’s fan runs about 900 rpm and is quiet. It doesn’t move a huge volume of air, but that’s part of the tradeoff for the noise reduction. Upon putting my ear closer to the fan, I notice a slight ticking sound and motor noise….nothing substantial, but it is definitely audible. When placed in a closed P180 case, however, the noise disappears. Overall, the Ninja’s fan is very quiet.

The Noctua’s fan runs at 1200 rpm and is virtually silent. It moves a higher volume of air than the Ninja and doesn’t exhibit any motor noise whatsoever. Putting my ear close to the fan results in the very faint “whoosh” of air movement. For all intents and purposes, the Noctua is absolutely silent, especially in a closed case. Clearly this is a superior fan in both regards, two critical criteria in trying to achieve cool and quiet.

The Zalman’s fan operates between 1300 and 2600 rpm, as it is adjustable. At low rpm’s the fan is quiet, but doesn’t move much air as the tradeoff. At its highest level, the airflow is respectable, but nowhere near the volume that is being pushed by the 120mm fans on the other two heatsinks. However, when at the highest airflow, the Zalman is very audible, even when the case is closed. It exhibits a higher-pitched “whir” and clearly does not achieve any semblance of quiet. This is a critical flaw: not only is it far louder than the other heatsinks, but it also does not have as high an airflow either. This severely cripples the Zalman.


TEMPERATURES
This is where the subjectivity ends and the facts speak for themselves. This is also where a heatsink’s true worthiness lies….if a CPU can’t be kept sufficiently cool, then beautiful and quiet simply does not matter one bit. Passing the quality and noise judgement is fine, but to be a true success, a heatsink must simply perform just as well, if not better.

As previously indicated, the closed P180 case was positioned near a furnace vent to increase ambient temperature. In all tests, the ambient temperature was maintained at 29 Celcius (85 Fahrenheit). This would reasonably approximate minimum summer indoor temperatures without adequate air conditioning, or winter temperatures near any heater/vent source. Again, to approximate real world conditions.

In all testing instances, the CPU was run both at stock 1.86 ghz speed and overclocked 3.20 ghz speed. Idle temperatures were recorded after booting into Windows and letting the system sit for 20 minutes. Load temperatures were recorded after running dual instances of Prime95 after 20 minutes. Highest idle and load temps were recorded with Intel’s TAT, and were cross-checked with SpeedFan’s readings. Differences of 15C Tcase and Tjunction (credit: CompuTronix’s Core2Duo Temperature Guide) of +/- 2C were considered sufficient for alternate verification. This was done for each heatsink. Note that all heatsinks were tested, then reseated on the CPU after proper cleaning and reinstallation of thermal paste, then tested again to ensure no singular anomalies skewed the results.

The results are as follows:





To be honest, the results were not what I had expected. Given the popularity of the Scythe Ninja as a “top quiet performer”, I was expecting better. As indicated, it was reseated and retested to ensure accuracy. The temperatures were the same for both test runs, within 1 degree Celcius.

The Zalman, with its 92mm fan, admirably kept pace with the Scythe, though it was far more audible at the highest airflow (I should mention these results were for the Zalman at the highest rpm to test maximum cooling ability).

Both the Zalman and Scythe flirted with unsafe temperature levels at load. However, the Noctua didn’t even cross into 60C territory, and actually managed to stay well enough below. It was clearly lower on all temperature ranges, at stock and overclocked, at idle and load.


THOUGHTS:
As most people are well aware, achieving high performance while maintaining a low-noise system that is cool is quite a feat. Many people spend a lot of hard-earned money and time, researching and building such a setup. And one area that is particularly sensitive to this fine balance is the CPU. First and foremost, it needs to stay cool....the cooler, the better. Low noise is certainly a bonus beyond that.

Of the three heatsinks tested, the Noctua is clearly the superior product in achieving cool and quiet. It has an excellent mounting system that is easy to install, occupies a smaller footprint than the other products, utilizes a near-silent fan, and clearly offers the best cooling across the temperature and speed spectrum. Build quality is top notch, and its design is both sleek and functional. It is not the least expensive heatsink on the market, but it is neither the most expensive. Given its moderate price and impressive performance, the Noctua NH-U12F is simply an exceptional product and a wise purchase towards achieving a cool and quiet computer system.


CONCLUSION:
Noctua NH-U12F is a Skyguy’s 2 Thumbs-Up Winner.
April 15, 2007 2:07:10 AM

Well, I'm not trying to "shove it" to anyone. I'm trying to be as objective and thorough as possible. Truth be known, I've owned a Zalman 9500 for quite some time, even replaced the fan to make it quieter. But after seeing these results for myself, I simply couldn't continue using it, knowing that a better option exists.
April 15, 2007 2:11:46 PM

Good reviews. just a few thoughts though

The Ninja's fan is supposed to be a 1200rpm fan (Assuming its the latest REV B version), you say yours was 900rpm

The CNPS9500, its the LED version right not the AM2 or AT

I agree about the mounting, backplates are the way to go for high end heatsinks.

Also another negative for the zalman is no AM2 compatibility you need the AM2 version which then means if you switch to 775 you need to buy a new one.

The Noctua NH-U12F is great I've got one, I can hear the fan on my computer at 1200 (just). But I use a fan controller and have it about 1000 which is perfect.

Price wise in the UK the ninja is probably the best value ~ £29 against £33 for the zalman and £37 for the noctua

I might have a look at the NH-U9F while I check out some low profile heatsink

Again Good work with this
Related resources
April 15, 2007 4:12:11 PM

Good points, thank you for the feedback.

Yes, my only "regret" is that I was unable to test more heatsinks. I really, REALLY wanted to test the Ultra 120 with the Noctua fan on it, to see how it does head-to-head with the NH-U12F. Unfortunately, lack of resources like I said :( 

I could buy one I suppose, but timing is becoming an issue. I think the building Superintendent will turn off the furnaces within a week and then I can't guarantee I'll get the same test conditions. I could use a space heater I suppose and try to maintain a constant ambient that is very close to the original test setup.
April 15, 2007 4:24:43 PM

First of all thank you for the review, the only problem I had with it is I think it would have been fairer to have tested the Zalman 9700 although I do understand the time and financial constraints. It just goes to show that there are a lot of quality products out there and people should really do their research before buying.
April 15, 2007 4:30:59 PM

I certainly wouldn't argue with you on that, you are quite correct. In fact, there are a number of heatsinks that I would LOVE to test, the 9700 being one of them. I'd be particularly interested in finding that "sweet spot" for noise where it's inaudible, then seeing how it performs at that range. There are better performers out there, to be sure, but they are generally alot more noisy too. However, the 9700 has a distinct advantage in that the fan speed is adjustable, so I'd love to see how it does against these. The 110mm fan is a big improvement, and I'm VERY curious to see just how well it does.

In fact, the Zalman is widely regarded as probably the most aesthetic.....everyone seems to love it in a windowed case (including myself). But the noise and 92mm fan on the 9500 crippled its results overall for cool and quiet. But the 9700 seems to fix these critical flaws, so I'd be very interested in checking its performance. If it overcomes those 2 issues, the 9700 could make Skyguy's Winner list ;) 

*hint Zalman, hint* ;)  LOL.
April 15, 2007 4:38:27 PM

I have had Zalman products in the past and I think they strike a good balance of silence and performance but you are right in that Zalman's are designed to install and that's it with other heatsinks you have the option of choosing your fan/fans and coming up with the best solution that suits your needs. I am very interested in the new sinks from Thermalright I think when those are finally released they will be beasts.
April 15, 2007 4:52:09 PM

Yea that's the one I've been looking at I'm just waiting for some reviews and for it to finally get released. Also hope it doesn't snap my board in half, lol. That thing is huge.
April 15, 2007 4:58:21 PM

That IFX gives me the shivers....it looks amazing. The great thing about the Thermalright, Noctua, etc is that you can use whatever 120mm fan you like. I prefer the Noctua fans for their silence and airflow, but the beauty is that someone else can put in a Silverstone FM121, for example, suck every drop of performance if they'd like.

The purpose of this review was for cool and QUIET, however ;) 

But I am hoping to add other products in the future if I am able. I am also looking at investing in a "noise meter" to measure actual decibels. I believe I can also easily ensure a consistent test setup for the higher ambient temps. So here's to hoping I can do more in the future.....
April 15, 2007 5:08:47 PM

Quote:
Apperantly you actually can buy it even though thermalright calls it a preview


I saw that to the other day when I visited their site, it's just a preorder and the don't even have a release date yet. Plus I would like to read a couple of reviews, at that price add fan or two and your approching the cost for a entry level watercooling setup. I just want to see if it's worth it.

I have never owned Noctua fans but I have read good things about them. Another brand I would like to check out is SilenX, their specs are incredible although I know some companies have a habit of being rather creative when it comes to their products ratings.
April 15, 2007 5:48:52 PM

So, after all of the work put into your sys to make it quiet how loud would you ay it is under full load, mid-load, and idle?

Try to get a good 5min listening on each setting.

Im getting a very similar sys soon, and im also focusing on noise reduction, and ive read dozens of reviews on these products, but no full-pc noise real world experiences.

*EDIT* oh, and see if you can get a comparison btwn the Noctua 120mm fans and the Silenx fans. Ive heard a lot of good about them (but the Nockies are sexier ;) 
April 15, 2007 6:01:33 PM

Actually, my system is very quiet at all times. I have all Noctua fans and they are virtually silent, and stay at 1200 rpms. So, whether my system is at idle, half, or full load, the fans stay constant, and therefore the noise stays constantly quiet ;)  The ONLY difference is that my temps are higher at load, which is obvious.

If you pick good fans and components to keep things fairly cool, then your system should stay quiet regardless of the load. If, however, someone decides to increase rpm's and fan speed, then noise will increase. But as I've mentioned, there is a balance to be found. In this instance, fan speeds stay constant and quiet, and temps rise, but still within acceptable ranges as shown by the Noctua heatsink results.

That's probably the best way to keep a system quiet, regardless of load.....just ensure you have good components to begin with. As my comparison shows, not all heatsinks are as quiet, or as good at keeping things cool. So it's absolutely CRITICAL to find products that achieve that balance, or else all that hard work will be in vain, because as soon as you push your system it will get loud or hot.....in effect wasting your time and money you've already invested.


And let's not go down the SilenX road......fans per se were not the purpose of this comparison. Besides, if you do a bit of digging here and at Silent PC Review, you'll find some very.....interesting.....information about SilenX. However, I won't comment any further than that in this thread.
April 16, 2007 4:37:38 AM

Quote:
And let's not go down the SilenX road


Sorry about that, must have hit a sore spot. I know the fans weren't the point of the review I just mentioned them because they look too good to be true (which means they probebly aren't).
April 20, 2007 7:36:56 PM

Considering the Noctua NH-U12F over the Tuniq Tower due to ease of installation. How do they compare in terms of cooling? And as far as vendors go is Xoxide dependable and probably the cheapest?
April 20, 2007 8:17:45 PM

Could I get a 3.4ghz or 3.6ghz oc out of the noctua?
Anonymous
a b K Overclocking
May 9, 2007 3:41:32 PM

Thanks for the well-written review, Skyguy. I had definately been leaning towards going with the Noctua for my upcoming build and now my mind is made up. I appreciate quiet over the slight decrease in temps that I could get from using some other heatsinks. (Especially considering that my current computer is a 7.5 year old Dell that constantly sounds like there is an airplane taking off from inside my house!!!)

-J.
May 10, 2007 4:16:10 PM

Yep, couldn't get that, wasn't released yet either. Put a quiet fan on it, and my guess is there's a new sheriff in town ;) 

I'm working on a much more complete list of heatsink comparisons with someone else.......no ETA on a writeup though, but it'll have all the heavyweights, including some newly released products as well.
May 13, 2007 2:59:04 PM

Quote:
Either the link is retarded or the site is down, I have a cable line and it's been 4 minutes with no sign of loading

But I did see a preview of it, and it looked pretty nice

Rugger's link worked fine for me.
@Skyguy: Great write up - looking forward to the expanded comparatives.
!