I'm really impressed that you guys are doing this, it shows great faith in your product. Given you're entering a rather saturated market, that's admirable.
I will not enter, as the liquid pro I'm using at the moment requires lapping each time one uses a different tim. I can't face lapping the cpu and waterblock again.
Good luck nonetheless
edit: If you'd be so kind, what is the thermal resistance you expect to be associated with the ihs : tim : heatsink interface? Assuming standard mounting pressure / tim applied as per instructions. It's really very difficult to find any data on this, and it would be useful to have a ballpark figure to work with.
Only data I have readily available - ICD Standard data point is the comparison. I have been familiar with liquid metals as something of a lab curiosity for the last 20 years. Gives you thinnest possible joint but they are awkward for the general user to implement. I believe Intel spec is flat and // to .02 on the IHS? This I believe cause some inconsistencies with some users with unlapped sink/IHS having to do with filling the gap on less than optimal contact.
I have only a couple of comparisons, as you say it is awkward to clean ( I use 000 fine steel wool on my test dies as it wicks into the wool although I do not recommend using on anybody's system as the metal fibers will play havoc on the electronics) so hard to get people to try ICD on a survey like this.
I have a metal pad and a liquid pro comparison from a German giveaway we did just before Christmas I posted on the results thread
here scroll down a bit. Inconclusive as the sample is small. I do believe the LM's work well when contact is good at low and high pressure but have a weakness if contact is light.
Common knowledge that the more liquid compounds have pump out reliability problems but I do not know whether this extends to the LM's as I never had experience or read anything on the reliability front.
In any event twenty years of thermal compound development have reduced the difference between using a solder joint to about 3 - 0.4 C above the die temp for the competitive compounds. Further development may reduce this difference by a few tenths of a degree, but for all practical purposes at the power levels we are operating at this is pretty much the end of the road. This is the brick wall and most likely in the near future the grouping will get tighter in this range with users maybe not focusing so much on performance but cost and reliability.
Would prefer IC24 to perihelion!
Perihelion is a top competitor in user tests to date it is only second to ICD on performance at much lower cost than ICD or comparative competition compounds. It has the same viscosity as ICD for long term reliability minimizing pump/bake out failures so do not discount the advantages on the Perihelion.
We have $100's of thousands in test equipment and so proceed with some confidence that we have great compounds but... that's a lab result and academic, the proving ground is/are what the end user sees on hid desktop is what counts.