Originally posted by: Eeyore69
I don't see a way of using plastic ties to secure the upper portion of the fan to the heatsink since the fan sticks above the heatsink. But I am thinking of re-bending the standard clips to fit or maybe just drilling additional clip mounting holes in the heatsink fins themselves positioning them farther inboard so the clip will clamp to the nearside of the fan. I'll have to think about it and decide which way to go.
I still have time since I need to lap the heatsink since it is not flat. It is highly polished but just not flat. And after I burn-in the cpu for awhile to make sure it doesn't crap out I will lap it also. I sure wish cpu manufacturers would do this since it would save me time. But then what else do I have to do but lap a cpu? And I do like DIY projects anyway.
_________________
Sorry about that, Eeyore69. I forgot about the 120mm fan-option. Even so, there are probably any number of fixes to that situation which still use wire-ties, although wire-ties aren't the only way to go.
_________________
THE LATEST REVIEW ON THE ULTIMA 90
_________________
Well, this is a slight surprise, since another review had actually shown performance which exceeded the Sunbeam Tuniq. Here, they show it neck-and-neck and tied, depending on fan choice.
I'm again wondering -- Eeyore69 and others may know firsthand -- if they continued with this convex-base design, when all of us have deplored that choice. If so, I'd expect the earlier review in comparison to the Tuniq to be vindicated, even if the test and results were somehow unscrupulously skewed in the Ultima's favor with that review. I was actually hoping that the Ultima would equal the performance of the U-120-Extreme, but what can you expect?
Looks like I'll be keeping my "Extreme" on the Quad-core, and choosing the Ultima 90 for use in a smaller case for a dual-core I'm building for another member of the family as a Xmas gift.
On the matter of the X6800 test-bench choice. It's clear to me that the over-clocked thermal dissipation for the C2Q doesn't scale like the C2D processor, but you could intuitively understand why that's the case. That is, beyond the stock TDP difference between 65W and 105W, the number of cores in the C2Q would have a non-linear impact, I would think.
But for comparing different coolers, ANY choice of current processor will do -- we only need a test-configuration that is consistent between all coolers tested. How that translates into "adequacy" for the C2Q is another matter, but those of us running the C2Q on air would agree that either the Ultima or the U-120-Extreme are more than adequate -- as is the Tuniq.
We'd still like to see advances in heatpipe-cooler performance such that idle-to-load spreads are reduced further, and such that temperatures at both idle and load are lower.