Horizontal vs Vertical - Some evidence?
 

Was just reading through Nexus' site, looking at some of their products when I noticed this page:
Nexus LOW-7000 R2 UNIVERSAL Ultra Quiet CPU Cooler, only 7cm high!

Specifically, these two images:
http://www.nexustek.nl/images/lowr2mountingok.jpg
http://www.nexustek.nl/images/lowr2mountingng.jpg


Now I scratched my head for a moment there, wondering why Nexus would restrict the mounting of this cooler with the heatpipes oriented perpendicular to the "VGA card slot".

Then I noticed this:

"Please follow these positioning instructions as much as possible when installing your LOW-7000 CPU cooler into a tower chassis PC Case. The correct positioning guarantees an optimal performance of the heatpipes.
When installing the LOW-7000 in a desktop or horizontally oriented pc/htpc case (where the cooler will be positioned with the fan horizontally) you can choose any positioning method."

I believe this to be the first definitive proof I've seen that heatpipe coolers do not always behave the same depending on their orientation (horiz. or vert.). I remember I mentioned this in a HSF review comment thread a while back, but it's interesting to see a manufacturer admitting to the differences. In the case above, it seems as if when using this cooler in a vertical (tower) type setup, the heatpipes would not work efficiently with the heatpipes being vertical as well.

Interesting food for thought - perhaps a reason to re-test some of the HSFs and why sometimes they perform better or worse without explanation. It seems to come down to its orientation and the orientation of the case. Discuss?

http://www.hardwarecanucks.com/forum...-review-7.html


;)

Wow, that article was months before I even joined the forum, lol. Slight fail on my part, but still worth discussing I'd say. Testing methodologies determine the performance of Heatsinks based on their orientation and positioning? Interesting!

I ma curious how well this holds up today. I ask because my own testing in the last 2 days of a venomous and 1283 xigmatek show almost no difference at all in fact it was so close that i would account this as mounting luck, within 1/2 a degree:thumb:

Heatpipe orientation dependence really depends on the design of the heatpipe. Cheaper heatpipes will work better in certain orientations, but better ones are basically unaffected by it. I saw an article somewhat recently where they cut open a bunch of heatpipes and showed the internal structure to demonstrate different types of heatpipes, but I can't seem to find it right now.

I remember that article you;re talking about, and I think there was one type of heatpipe that had grooves in it and worked better oriented with the groves facing upwards I believe.

edit- here it is:Heatpipe Wick Structures Exposed: Sintered, Groove and Mesh - FrostyTech.com

Heh... I spent quite some time scratching my head over the performance of that particular cooler when compared to others which on paper looked like they should perform on par.

Seems to be the low profile coolers which are affected most by this, maybe it's because the heatpipes have more of their length in the curved portion.

Or Perhaps because the low profile coolers have their heatpipes "flat" in relation to the motherboard? Im beginning to wonder if coolers should more often be oriented N/S so that no one "side" of the heatpipes are lower than the other to reduce its efficiency.

It's certainly information that we'd benefit from if the manufacturers would provide their testing results WRT most efficient orientation as was the case in your original post.

That's the one.