by Sergey Lepilov
02/06/2008 | 03:06 PM
Every human being is unique. We all subconsciously strive to stand out among others. There are multiple ways how this goal can be achieved. Some of us are extremely smart, some are unearthly beautiful, some are kind and helpful, some possess other respectful traits of human character. Those, however, who feel like this is not enough usually resort to some exclusive things, like limited edition cars, watches, “haute couture” outfits, jewelry, etc. Of course, in the computer field there are much fewer things like that. We can certainly recall 1000-piece batch of exclusive Albatron PX845PEV-800 mainboards, 500-piece batch of limited edition ATI Radeon X1950 XTX Uber Edition graphics cards, Dell XPS 600 Renegade gaming system. Over time things like that turn legendary, and it is very likely to happen to Scythe Ninja Copper cooler launched by Scythe’s 5-year anniversary.
People started talking about the fashionable copper Scythe Ninja modification right after the company launched their original aluminum Scythe Ninja, and Ninja Plus a little later. Back in 2005 they demonstrated the copper modification of this top cooling solution at Computex 2005, but it never really went any further: the cooler never made it to mass production. In the meantime, Japanese company’s fans and dedicated fans of the Scythe Ninja cooler didn’t give up and kept expressing their hopes about the copper fellow in numerous hardware forums. Of course, their expectations were absolutely natural, because the aluminum Scythe Ninja was the best air-cooler at that time, and the efficiency of a copper modification promised to be absolutely stunning.
However, times passed, cooler makers designed and released new products that already started to outperform the Scythe Ninja (Scythe actually launched their Infinity cooler as well), and the copper Ninja was nowhere to be seen. The stimulus for the Copper Ninja to be finally released was Scythe’s 5-year anniversary, so they decided to commemorate it with a limited edition of new Ninja coolers featuring copper heatsink rib array instead of an aluminum one. Experienced overclockers may be a little skeptical about replacing aluminum heatsink with a copper one. Besides, I also wouldn’t expect a tremendous cooling efficiency boost from that. Nevertheless, Ninja Copper is indeed unique, and a few other interesting innovations made in it are totally worth our closest attention.
Slim-looking, bright-colored box stresses the uniqueness of this cooling solution:
The front of the box bears a full-size cooler photo and a few icons indicating that Ninja Copper supports all new CPUs, can work in passive fanless mode and comes with a silent 120mm fan. A separate tag on the front of the box states that this cooler belongs to a limited anniversary edition of Scythe products, because as I have already mentioned earlier, this company has recently turned 5 years old.
The other sides of the box are also very informative:
When you open the box, the 120mm fan sits on top, then comes an unusually heavy copper heatsink, and at the very bottom there is a small flat box with accessories. Everything you need to install this cooler is inside:
According to the info on the box, Scythe Ninja Copper is made in Taiwan. As far as we know today, the cooler has been manufactured in limited quantity of only 1000 pieces by the company’s anniversary. Looks like Scythe Ninja Copper will not go into real mass production.
Of course, when you take Scythe Ninja Copper out of the box the first impression is: “oh god, how heavy it is!”. Well, over 1kg (1015g) of net heatsink weight is a rare thing these days. Please understand me correctly, I am not trying to scare you, because mainboard textolite is designed to withstand even more severe pressure and loads. But the new cooler is so heavy that it is hard to simply get it out of the box. The best way to do it is to turn the box upside down and let it slide out, just watch your feet, otherwise you may find out what a kilogram of copper feels like when it falls :)
Just look at this beauty:
Nothing has changed dramatically in Scythe Ninja Copper design, but they certainly made it look much more luxurious by simply replacing the aluminum heatsink array with a solid copper one.
The newcomer is 110mm x 110mm x 150mm big and is based on six copper heatpipes placed in two layers, one on top of the other:
The heatpipes carry 23 copper plates, each about 0.3mm thick, with a 5mm distance between them. It’s been a while since we last tested a cooler featuring an array of copper heatsink plates of that size.
Note that one of the differences between Scythe Ninja Copper and its predecessor is the heatpipes spread out at a distance from one another:
As you remember from our previous review, the heatpipes of regular Scythe Ninja cooler were put together in groups of three. The new cooler has its heatpipes spaced out from one another, which allows not only to more evenly distribute the heat over heatsink plates, but also to reduce the airflow resistance. Strange that Scythe engineers needs over two and a half years to figure this out, because until this day Scythe Ninja was only available with heatpipes set in close groups of three.
The ends of all heatpipes are topped with aluminum caps. So, if you look at the cooler from above, you may get the impression that it features 12 and not 6 heatpipes:
There is an aluminum heatsink at the bottom of the cooler. It is of a little bit different shape than the one of a regular Scythe Ninja:
You cannot remove the heatsink from in-between the heatpipes, but you can unscrew it from the cooler base and lift up a little bit to see that there is no thermal interface of any kind between the heatsink and the top row of heatpipes:
It’s a real pity that a $70+ cooler has an upsetting flaw like that, which we can only see in budget cooling solutions most of the time, and not in the one that tends to be the top choice.
I used a cotton Q-tip to apply some thermal interface to the heatpipes beneath the heatsink:
Then I screwed the heatsink back on and removed it again. The thermal grease imprint on the heatsink base indicated that even if there was a layer of thermal interface between them, the heatsink had very little contact with the heatpipes - only a few thin lines, and not the entire surface as I wished. I wonder why they would need this aluminum heatsink at all, if it doesn’t get in contact with the heatpipes but simply hangs on two retention plates. I don’t think that applying more thermal grease would solve the problem. It would be great if the manufacturer could pay more attention to this issue, if Scythe Ninja Copper ever goes into mass production.
The cooler base if protected against scratches with plastic film sticker that bears a large exclamation mark and a warning to remove the film before installation:
The cooler base is made of solid copper, but is covered with a thin layer of nickel alloy. The base finish quality and evenness are absolutely stunning:
Another distinguishing feature of the new Scythe Ninja Copper is an improved fan from the Scythe Slip Stream 120 series:
The new fan series features smaller rotor diameter and bigger operational surface of the fan blades. SY1225SL12L model that comes with Scythe Ninja Copper rotates at ~800RPM generating 40.17CFM airflow. In this case the level of generated noise doesn’t exceed 11dBA. Sounds very impressive, doesn’t it?
It is very nice to see that there were multiple changes made to the new Scythe Ninja Copper and that the manufacturer didn’t simply replace aluminum plate array with copper. Now we just need to check it out in action. But before we move on to the tests, let’s say a few words about the installation procedure.
Scythe Ninja Copper may be installed onto Socket 478, LGA775 and Socket 754/939/940/AM2 mainboards. The latter one requires a special retention bracket with a lip that needs to be fastened to the cooler base first:
The clip catches onto the standard plastic socket frame that is why Scythe Ninja Copper may only be positioned facing one of the two possible directions. To be fair I have to say that it doesn’t really matter for Scythe Ninja Copper, which way it is facing, because its heatsink is symmetrical. The only thing that comes to mind is directing the airflow through the lower “fake” heatsink, not that it would actually matter that much.
In case of an Intel platform, the installation procedure becomes a little bit more complicated. At first you need to install two retention brackets on the mainboard PCB using bundled screws:
A backplate with soft padding should go on the other side of the PCB:
I have to say that it is very inconvenient to tighten the retention screws, because you have to hold the board, the backplate and retention brackets in place at the same time.
A completely different retention bracket needs to be fastened to the cooler itself:
And then you slide the retention hooks into the slits on the brackets and press the cooler against the CPU until the hooks on the opposite side catch on to the slits, too:
You should keep in mind that this is extremely hard to accomplish inside a system case. The clips are beneath the heatsink, so you will have really hard times trying to get to them when the mainboard is already installed. However, in an open testbed, the whole thing will just take a minute.
Scythe Ninja Copper looks like this once installed into a system case:
Note that I had to remove the memory module from the first and the third DIMM slot in order to install the fan the way it is shown on the photo above. Of course, you can affix the fan on the side facing the mainboard chipset, but what if the chipset heatsink is too tall? You can also put the fan higher up, but in this case it will not cool down the components in around-the-socket area, which is also important for contemporary systems, and especially during quad-core processor overclocking.
Technical specifications and recommended price of the newcomer are given in the table below:
Scythe Ninja Copper cooler and two of its today’s competitors were tested on an open testbed as well as in a closed system case with the following configuration:
All tests were performed in Windows XP Professional Edition Service Pack 2. SpeedFan 4.34 Beta 38 was used to monitor the temperature of the CPU, reading it from the CPU core sensor:
Its readings matched those from the Core Temp 0.96 utility. The mainboard’s automatic fan speed management system was disabled for the time of the tests in the mainboard BIOS. The CPU thermal throttling was controlled with RightMark CPU Clock Utility version 2.30:
The CPU was heated up with OCCT (OverClock Checking Tool) version 1.1.1b in a 23-minute test with maximum CPU utilization, during which the system remained idle in the first and last 4 minutes.
I performed at least two cycles of tests and waited for approximately 20 minutes for the temperature inside the system case to stabilize during each test cycle. The stabilization period in an open testbed with the mainboard in horizontal and coolers in a vertical position took about half the time. The maximum temperature of the hottest CPU core out of the two/four in the two test cycles was taken as the final result (if the difference was no bigger than 1°C – otherwise the test was performed at least once again). Despite the stabilization period, the result of the second cycle was usually 0.5-1°C higher.
The noise level of each cooler was measured according to our traditional method described in the previous articles with the help of an electronic noise meter – CENTER-321. The subjectively comfortable noise level of 36dBA is marked with a dotted line in the diagram; the ambient noise from the system case, without the CPU cooler, didn’t exceed 33.4dBA when measured at 1m distance.
The ambient temperature was checked with an electronic thermometer that allows monitoring the temperature changes over the past 6 hours. During our test session room temperatures stabilized at around 25°C. It is used as a staring point on the diagrams. Note that the fan rotation speeds as shown in the diagrams are the average readings reported by SpeedFan, and not the official claimed fan specifications.
It is quite logical that we decided to compare Scythe Ninja Copper against the regular Scythe Ninja (Plus) cooler with aluminum heatsink plate array. Unfortunately, we didn’t have the regular Ninja in the lab at the time of tests. However, it was really no problem at all, as we had a very “similar” OCZ Vindicator cooler, which will be one of the today’s testing participants, too. Besides, it will also be interesting to see how the new Scythe Ninja Copper performs against the today’s air-cooling solutions leader – Thermalright Ultra-120 eXtreme cooler, which actually costs about the same as the anniversary edition solution from Scythe.
All coolers were tested with two types of fans. First of all, each of the participants was tested with Scythe Ninja Copper fan – the Scythe Slip Stream 120 working at ~860RPM (according to the monitoring report). After that we tested the coolers with Noctua NF-P12 fan from the Austrian manufacturer. It was rotating at ~1380RPM. During the last test cycle the coolers worked with a pair of Noctua fans each: one installed for air intake and another one – for exhaust:
Now we are all armed and ready, so let’s get down to the results!
Intel Core 2 Duo E6750 CPU with G0 core stepping overclocked to 3.76GHz inside a closed system case with the weakest cooler of the three participating today, which is actually very close to its frequency maximum obtained with air cooling only. We had to increase its Vcore in the mainboard BIOS from the nominal 1.35V to 1.625V:
According to CPU-Z, Everest and SpeedFan, the voltage varied around 1.6V, and dropped to ~1.58V under high workload during our tests. Moreover, two processor tests from the OCCT suite launched one after another were passed successfully both times without any indication of attempted thermal throttling:
Let’s check out the temperature readings for the dual-core processor on the following diagram:
Well, the results are pretty interesting. First, Scythe Ninja Copper is 3-4ºC more efficient than its aluminum brother, which is not bad already. Second, the new Scythe fan proved to be great fit for the new cooler, because a more powerful Noctua fan has minimal effect on the cooling efficiency. It is not only the copper plates, but also greater distance between them that matter more in this case (although the gaps between the heatsink plates of OCZ Vindicator cooler are about the same). Third, Scythe Ninja Copper can even compete with Thermalright Ultra-120 eXtreme, although only in case they both have Scythe Slip Stream 120 fan. As soon as Ultra-120 eXtreme receives a more powerful Noctua fan, or two of those, it wins an indisputable victory, as we have expected.
Of course, we couldn’t help checking out the performance of the new Scythe Ninja Copper in passive mode, i.e. without a fan. The test was performed in an open testbed only, because there are other case fans inside a system case, which could make our experiment not 100% fair. As a result, dual-core processor with a passive cooler on it worked stably at 3598MHz frequency with the Vcore increased to 1.45V. The peak temperature reached 91ºC, all plates including the top one were very hot to the touch (I couldn’t stand more than a second), but the CPU didn’t go into thermal throttling and kept running at its maximum speed. Further frequency or voltage increase caused Scythe Ninja Copper to fail. However, the achieved result is already impressive enough, I assume. Especially since OCZ Vindicator as well as Thermalright Ultra-120 eXtreme also tested today couldn’t repeat the success of Scythe Ninja Copper in fanless mode.
Quad-core Intel Core 2 Quad Q6600 processor (B3 core stepping) with polished off heat-spreader cover overclocked to 3312MHz inside a system case when cooled with the weakest cooler of the three. Its 120mm fan was running at 860RPM. The test was performed with the processor core voltage increased to 1.45V:
CPU-Z as well as other above mentioned monitoring utilities reported a slightly lower actual voltage than what was set in the mainboard BIOS, namely around 1.425V. The OCCT test was once again passed successfully without any errors or throttling activation:
The results demonstrated by our coolers on a quad-core processor are given below:
Frankly speaking, I expected Scythe Ninja Copper to look as impressive on the hot Quad processor with B3 core stepping as it did on a dual-core one. However, I was wrong and I am actually very happy about it. To be fair, I have to say that the new Scythe Slip Stream 120 fan is not enough for Thermalright Ultra-120 eXtreme, because the gap between its heatsink plates is much smaller than by Ninja and its clone. Replacing Scythe’s fan with a Noctua one or two of those buys us about 1ºC or 2ºC respectively (we are talking about an open testbed only). At the same time, the same modification performed on Thermalright Ultra-120 eXtreme cooler improves its efficiency by 4ºC at first and then by another 2ºC, which makes this cooling solution an indisputable leader.
Compared with the results obtained on a dual-core platform, Scythe Ninja Copper running in fanless mode on top of a quad-core CPU performed a little bit more modest. The CPU worked stably at 3012MHz frequency (+25.5% above the nominal) with 1.3V core voltage (1.285V is the nominal). However, you should take into account that our quad-core processor is the one of the older B3 stepping, which is known for its hot “temper”. The chances that you will get one of these are getting smaller day by day, and the CPU with new G0 core stepping feature much lower heat dissipation, so their overclocking results with the Scythe Ninja Copper cooler in passive mode are very likely to be better. In conclusion I would only like to add that two other testing participants failed to ensure proper passive cooling of the overclocked Quad CPU.
Since all our today’s testing participants were tested with two types of 120mm fans and their rotation speed is far from being high enough to cause cooler heatsink plates to resonate, our test session turned into checking out the fans noisiness.
Here are the obtained results:
Of course, Scythe Slip Stream 120 fan that comes with Scythe Ninja Copper cooler is extremely quiet. It would be incorrect to call it completely noiseless, but it is actually really close to that. Higher-performance Noctua fan is a little louder, however our tests revealed, that installing this fan or two fans (for intake and exhaust) instead of the Scythe’s own has very little effect on the overall cooler efficiency.
No doubt, that Scythe Ninja Copper is a remarkable cooler. It would be a perfect choice for those who love silent systems, but could also please overclocking fans with its performance in active mode (with a fan). Stylish exterior and universal design could also be a great advantage for many potential users. But it only something that could happen, because Scythe Ninja Copper is very unlikely to make it into mass production, so its will probably end up in the systems of those close to Scythe or their friends. The price of the cooler is secondary and matters only from theoretical standpoint, even if a few coolers make it to retail. A legend is a legend.
What we would really want to believe is that the constructive changes made to the Scythe Ninja Copper such as spaced out heatpipes, more efficient fan bundled with the cooler, mirror-shining base surface, etc. will eventually find their way into the regular Scythe Ninja with aluminum heatsink. Moreover, we would like to see the lower aluminum heatsink to be fastened in a little bit better way, and maybe even modified for higher efficiency. And of course, it would be nice to see Scythe engineers improve the installation procedure for LGA775 mainboards, so that it could be simpler and more convenient to perform.
In conclusion I would like to list all the highs and lows of the new cooler model once again:
* - I didn’t mention high price among the lows, because solutions like that are never cheap.