IntroductionScythe Co. is a new entry into the very exclusive, very discriminating PC enthusiast market and they make an honest and successful attempt in many respects.
In the Box
The first thing you will notice when you take the heat sink out of the box is its incredible weight, due largely to the huge copper heat spreader. The construction of the NCU-1000 is very impressive, the fins are strong and rigid, the plates and screws are all finely machined.
Upon closer inspection of the fin unit you will notice that the body is comprised of many parallel tubes running down the side, along the bottom and up the other side of the unit. The fins are attached in wave formation along the inside and the outside of the U-shape. The tubes are filled with thermally conductive liquid and essentially act as heat pipes, transferring the heated liquid to the top of the unit an the cooled liquid back down to the bottom in a cyclic fashion.
Actually there is a nice little explanation on the box:
" The Heatlane is TS Heatronics's original technology, which circulates the working fluid within the meandering capillary tube. The technology based on the operative principle which is completely different from the conventional heat pipe. It realizes effective heat transfer not only horizontally but also from the top to the bottom (top heat) as well as better heat transfer capability than the conventional heat pipe."
Because the NCU-1000 is very heavy, it is necessary to take advantage of the 4 mounting holes on the Pentium 4 motherboard, instead of using the standard plastic mounting bracket. To install the NCU-1000 the motherboard must be removed from the case and the support plate and insulation pad sandwiched on the back with the mounting screws coming through to connect with the copper heat spreader on the front of the board. The fin unit is attached to the heat spreader via the presser plate and the 4 presser plate screws. Finally everything is clamped down on both sides of the board and held firmly (yet gently) against your P4.
This is definitely a ginormous heat sink, weighing in at around 1.5 pounds.
The instructions indicate that the NCU-1000 should be mounted with the airflow vertical, I believe this to be so that gravity does not adversely bias one side of the heat sink over the other. With airflow vertical, the liquid in the tubes is always at a constant height.
Now comes the fun part, the disclaimer:
-Use this product at your own discretion.
We assume no liability and provide no warranty for any mechanical damage of parts including a motherboard, CPU, memory, graphics board, other cards, hard drive, CD-ROM drive, and power supply unit; and for any software damage including data and programs in the hard drive.
-This product has been manufactured for a general consumer. If you want to use it for a business application, contact our company for instructions.
-This product is designed with maximum CPU load for P4 up to 2.8GHz as e-mailing, internet browsing, using word processor, spreadsheet software or equivalent at ambient temperature up to 25C.
We assume no liability and provide no warranty for accidents and damages caused by the use of the product at higher CPU load environment than the above.
This product does not guarantee performance at high CPU load environment such as data encoding, decoding, benchmark software running, etc.
-Said performance of this product is according to the measurement value at our own test environment. The value may be different at other environment and/or conditions.
-This product does not guarantee CPU overclocking. TS Heatronics shall not be liable and assumes no responsibility whatsoever for accidents and damages caused by the use of the product for an unspecified operation, such as overclocking.
-There may be some products with slight flaws that were caused during the manufacturing process. Such flaws, however, will have no adverse effect on the product's performance.
Scythe Co. was very adamant about getting the point across to me, they even went as far as printing the warning out from the webpage and circling it in red ink with the note "VERY IMPORTANT!!". So let it be said that while Scythe Co. stands behind the NCU-1000, they do so with a large disclaimer and long list of guidelines.
With that, on to testing!
While there are usually no great expectations from passive coolers, I was optimistic with the NCU-1000 because of its unique technology, huge size and the claims from its manufacturer: TS Heatronics. As it turns out, Scythe Co. is not the company that makes the NCU-1000, they merely distribute it.
The manufacturer's claims follow:
These are the system specs and ambient temperatures:
CPU : Intel Pentium 4 2.8GHz
Memory : DDR-SDRAM 512MB
Motherboard : GIGABYTE GA-8GEM667K
OS : Windows ME
Thermometric software : MBM5
Outside Ambient Temperature (OAT) : 20C
In-case temperature : 24C (before bench marking, measured by sensor)
Power Supply : AOpen Strong Power 400W
PC Case fan : Nil
These are my results:
And these are my system specs and ambient temperatures:
CPU : Intel Pentium 4 2.53GHz
Memory : DDR-SDRAM 1024MB
Motherboard : ABIT IT7-MAX2 v2.0
OS : Windows XP Pro
Thermometric software : MBM5
Outside Ambient Temperature (OAT) : 23C
In-case temperature : performed on workbench
Power Supply : Antec TruePower 430W
PC Case fan : Thermaltake case fan, 32CFM (for active test)
As you can see there appears to be some discrepancy. My ambient temperature is colder (air around their heat sink was 24C, air around mine was 23C) and my CPU dissipates less thermal power (6.9W less), but my measure temperatures are significantly higher. While there are numerous factors that could account for a small degree difference, such as thermal paste (I used Arctic Silver 3), poor mounting against the CPU (I checked and there was full contact) and inaccurate thermal sensors, I feel that none of these could account for the margin between my results and the claimed results. I believe a possible cause of the discrepancy could be a difference in airflow through and around the heat sink in our varied testing environments, because as I show later the NCU-1000 works much better actively cooled.
It was just pointed out to me that I am an idiot. I was not careful in my examination of the manufacturer's claimed results so I did not notice that the temperatures on the graph are adjusted to not include the ambient temperature. That explains the exceptionally low temperatures. I have updated my results graph to more accurately reflect the relationship between my temperatures and TS Heatronics's temperatures. I appologize for any undue conclusions I came to and for any confusion I caused.
When I first turned on the computer, the initial temperature recorded by Motherboard Monitor 188.8.131.52 was 45C, as I let the computer sit under 0% CPU load the temperature gradually began to rise during a 45 minute period up to a steady stopping point of 55C. I can only assume that the working fluid could not circulate fast enough to release the heat off into the environment so it became saturated, only finally reaching equilibrium at 55C. It must be noted that there was negligible airflow in my testing environment, which could partially explain my higher temperatures. While under Prime95 load, the temperatures shot up so quickly that I feared for my P4 and stopped the test at 60C.
After passive tests I decided to give the NCU-1000 a whirl while actively cooled to see if it performed any better. I used an 80mm Thermaltake case fan that puts out 32CFM at 2050rpm. The results indicate that it performed better than while passively cooled, as to be expected. From what I remember, these numbers seem about on par with how well my P4's retail heat sink performed. With a good high CFM 120mm fan and some creativity you might have a half decent HSF unit.
With the passive tests not producing favorable results, I can only recommend the NCU-1000 if you have decent airflow in your computer case. Even then though I would have to only recommend the NCU-1000 to those who do nothing more than light computer work.
Perhaps the most interesting thing about the NCU-1000 is that the "working fluid" actually worked very well. While it did not cool the CPU effectively while passive, it did transfer heat to the top of the heat sink very well, in fact the top of the heat sink was hot to the touch. This is not often the case with regular heat sinks, as the heat does not transfer well radially throughout the sink.
I do not doubt that a heat sink fan unit with the Heatlane would work very well to spread out the heat and cool the processor, so I look forward to such a product geared for the PC market in the future.
The innovative Heatlane technology was very intriguing and most impressive
in its operation, however its implementation in the NCU-1000 form left something
to be desired. For Scythe Co and TS Heatronics' cool technology but unimpressive
performance I give the NCU-1000 a 6/10.
Reviews, Articles, News, All Reviews...
Reviews, Articles, News...
Career Center, News Users, Login...
A Guru's World, CPU/Memory Watch, SLDeals...
Forums, Register(Free), Todays Discussions...
Search, About Us, Advertise...