EK Supreme HF Orientation & Performance

Yes, I know that horizontal orientation (flow from left to right) is the best one for performance.
However, I am considering something different due to tube routing.
I can swear I've seen a thread with someone testing EK Supreme HF in all 4 orientations and posting the results. I can not find it, though, and I spent the last 2 hours looking for it.
Mostly interested in the performance of the block rotated by 90 degrees counter-clockwise (flow from bottom to top, should be pretty restrictive, and I recall like 0.5C performance hit), and 180 degrees (although I assume the latter should have no performance hit).
Feel free to help me find that thread (or review), or post your experiences. Thanks! :)

Sandy Bridge die has a obvious rectangular- and not square shape as the Ci7 predecessor. Maybe you would see a tiny gain under load if you would mount the waterblock rotated 90° (so the EK branding is rotated 90°). It's hard to say - if the block is mounted in a normal position the die would be covered with more channel. If the block is rotated 90° then the length of channels passing over the die is longer. All in all there shouldn't be a measurable difference.

http://www.lifeatmost.com/wp-content...ridge-chip.jpg

Like this:
http://www.dazmode.com/_Reviews/CPU2...0FC%206%20.jpg

Regards,
Niko

as long as the flow is correct the position the block is in doesn't really matter if i recall correctly.

tiborrr: do you happen to know, how lga2011 hexa core cpu die is supposed to look and what orientation might be preferred for them? Similar to SB?

Good point Churchy, not all of us have upgraded : )

heck for that matter.. what about AMD's?

Thanks for your replies! :)

Quote:

---

Originally Posted by tiborrr

Like this:
http://www.dazmode.com/_Reviews/CPU2...0FC%206%20.jpg

---

Yep, that's the one I'm talking about (90 degrees counter-clockwise).

Quote:

---

Originally Posted by tiborrr

Sandy Bridge die has a obvious rectangular- and not square shape as the Ci7 predecessor. Maybe you would see a tiny gain under load if you would mount the waterblock rotated 90° (so the EK branding is rotated 90°). It's hard to say - if the block is mounted in a normal position the die would be covered with more channel. If the block is rotated 90° then the length of channels passing over the die is longer. All in all there shouldn't be a measurable difference.

---

I actually don't have SB, was hoping to sort out my i7 920 rig.
Either way, I am correct thinking that there should be little to no difference, right? :)

Correct. Just rotate the waterblock in the direction which allows you the easiest tubing routing! :up:

@churchy: Don't know yet, haven't seen the die photos yet.

I actually managed to :banana::banana::banana::banana: up and reversed the flow direction as well as rotating the mount (mount change was intentional to test a new tubing run). It was only 2c hotter with the flow direction reversed through the block and the 90 degree rotation.

As Tiborr guessed the difference is negligible.

When you guys say, Left to Right does that not depend on the way the chip/socket sits? What I mean is that in my DFI the chip bottom is at the top (of the socket) - so should I turn the block to run Right to Left?
http://www.xtremesystems.org/forums/...pictureid=1133

At present I have my LT block running Top to Bottom but I am replacing it with a Full Copper HF this week - does the new "Easy Mount" mean I don't need the EK Back-plate (s1366) or can I still choose to use the "old" mounting? Just hope it has plate #1 installed as I do not fancy opening a new block...........