Isaac,
My rig will not be entirely passive. Still, it will be more passive than the typical midtower setup housing a 120mm PSU intake above and a 120mm exhuast behind the CPU cooler. There will be only one 120mm exhaust behind my tower cooler, the top quarter of which will be ~1" directly behind it. It will also be in a SFF case which is lacking decent airflow atm.
However, if I understand your remarks correctly, use of the Ninja in a vertical orientation would be more effective than in a horizontal one due to the gravity-assisted fallback of the condensed heatpipe water (assuming otherwise identical setups). It appears my designs live on!
There has to be a point at which an increasingly-smaller amount of airflow is no longer enough to outweigh the benefits of convection airflow, as small as they may be, in a given setup. It doesn't appear that I will come close to this point, but it would make for a damn interesting read if someone were to test for this.
BrianE,
Rotating the NCU-2005 90 degrees as the seperate copper base allows still provides more convection cooling because the fins are still vertically oriented when the NCU-2005 is mounted horizontally. The fins may not be facing the ideal way to receive fan airflow in the review setup but they are still positioned correctly to benefit more from convection cooling than they would if the entire tower was positioned vertically.
You are right in the fact that I missed important information in that review. The real juicy info, however, is the results where the motherboard was placed horizontally (thus making the NCU-2005 vertical) and the performance was "virtually unchanged."
Guys,
The real question now is whether the heatlane tech of the NCU-2005 can benefit as much from the vertical orientation as the heatpipe tech of the Ninja.
The
marketing of the heatlane tech claims that "heat transfer capability is not extremely affected by its applied position," and that it is "Hardly . . . affected by . . . gravity." That's great for those who will be using it the traditional horizontal way, but this implies only marginally improved performance for those who will be using it vertically.
Take this heatlane pic and rotate it 90 degrees counter-clockwise to see what we'd be working with:
Thoughts?