I dissipate thermals on processors quite often.
Lol, that is possibly the best line I've heard in a while. Though .. do you? I mean,
you? Not your heatsinks?
And I have done a fair share of testing.
Cool.
Both of our points are actually valid, but only to a certain extent.
So here's how I will portray plate differences.
Which water block do you think works best of the 2?
Absolutely no idea, as I'm looking at them from the outside, and can't see pretty much anything about them. Utterly and completely impossible to tell.
Which cold plate works best, I have 4 shown here.
Depends on the chip they're covering as well as the cooler being placed on them (to some extent - as long as the cooler has sufficient contact and covers the entire reasonable IHS area, it's not very relevant).
The deal is, I'm probably the only one here that knows which cold plate would work best for high thermally challenged situations, I'll give a hint, it's not the Xbox 360 IHS plate, that's for sure.
... but a cold plate isn't an IHS. As for which is best,
obviously you know that - they're yours, and you have tested them. And obviously nobody else does, as it's a random assortment of metal bits that you've taken a photo of, that the rest of us can barely make a guess at from the photo. Is this supposed to prove anything beyond you having experience with these specific pieces of metal and the rest of us not having that?
If it's a small, dense chip like a Ryzen 5000/7000? Then they're most likely roughly equal, including the Xbox 360 one - assuming they're all flat, they all make perfect contact with both the chip and the cooler, etc. Of course, unless you're machining your own cold plates with high precision metalworking tools, you're not actually controlling that, meaning you're introducing all kinds of variability just from differences in flatness, stiffness, etc. Whichever ones are thicker would impede thermal transfer more; whichever ones are smaller would have a marginal detrimental effect on cooling from lower area (assuming whatever you have cooling them can actually cool the entire area of the plate), but as the
vast majority of heat from a small heat source is transferred as straight through the IHS as is possible, the difference from this will be minimal. On the other hand, if you're doing something extreme that can make use of the massive surface area of one of those giant metal plates, then they are given an inherent advantage because they're being cooled below ambient, making their heat soaking ability into an actual cooling ability due to the inherently high thermal deltas from that.
Heat storing into a system made for cooling is not good.
Thick cold plate, thick TIM layer, thick IHS are sub-optimal for removing heat.
Why a lot of air cooler manufacturers chose the direct heatpipe contact? Because they wanted to lower cost from the cooler by ditching a chunk of metal?
No, because it was in the way to heat pipes.
If a water block has a thick-er cold plate from a another water block its because it needs it to sustain form under mount pressure. Even 0.001mm deformation on the cold plate surface under pressure can hurt its performance. Which one can you bend easier? A bigger or a smaller surface having same thickness? So bigger coldplate surfaced blocks need to be thicker.
When it comes to heat transfer the thinner the better.
Thickness works as insulation even on materials with high heat transfer ability.
If a bigger, thicker coldplated block has better performance is due to effective surface on the inside and the amount of water that can let through it. Those two (inside factors) can compensate the thickness of the coldplate by preserving the inside fins colder than the small block. So heat transfer is grater, as long as there is a big radiator to keep water colder and closer to ambient (than the small block).
The only thing better on a cooling system that stores energy is the time factor that its needed to reach its heat soaked condition, nothing else is better.
Exactly this. All materials are thermal insulators - some are just (much) worse than others. Copper is a very bad insulator. Heatpipes are
horrible insulators. You want as little of said insulation as possible between your heat source and whatever is dissipating your heat into ambient air.
There is one relatively minor exception to this: with direct contact heatpipe coolers, you need a sufficiently thick IHS to spread the heat out to the sides so that it reaches as many heatpipes as possible. The reason this is a minor exception is that "sufficiently thick" in this case really isn't a lot at all, and any structurally sound IHS able to handle cooler mounting pressure will be sufficiently thick for this.
I know that what I'm saying on this post seems like in opposition with what I said about the 7000 IHS. But its not. The reason of thickness of 7000 IHS is the gaps that has around it and not its surface size.
I don't think you're right about this though. I'm quite convinced that the only reason for the Ryzen 7000 IHS is maintaining cooler compatibility with AM4 coolers. There's nothing else they could have made thicker - making the socket taller would have been hell with all those pins and their spring force; making the substrate thicker would be expensive and wasteful. There isn't anything else to change to make up the height difference between AM5 with a thinner IHS and the bottom of an in-spec AM4 cooler - so they thickened the IHS to fill the gap, to maintain the same Z-height between the two.
Copper - better heat conductivity than aluminium, also bigger surface area = better heat dissipation
None of those are aluminium - silvery colored IHSes are nickel plated copper.
).
