Hey there, happy to discuss this, want to keep it casual and fun like liquid cooling should be
Much of this is based on our experience. For general reference of where we're coming from, we have a background in aerospace, hydraulic and medical manufacturing with some fancy engineering degrees.
Also, I thought it was general knowledge that more surface area is better for cooling. Bigger air coolers are better than small air coolers. Same with waterblocks. Same with radiators. Same with any kind of cooling. The entire world of cooling works with cooling like this -- PCs, cars, air conditioners, etc.
Maybe I'm missing something your points or not understanding what's being disputed here? Any competing products you can point to that illuminate what you're talking about?
Actually yes, retaining BTU before dissipation is exactly how it works. BTU doesn't transfer through any material instantly, there's a set time limit. ...he time it takes to heat a material up, is not always reflected by the amount of time it takes to cool off.
The heat retention of a block of copper will be reached nearly immediately. A cold plate could be an inch thick, that CPU will shut itself off instantly. That's why it has to be cooled, right? More surface area cools faster.
As for the idea that holding more BTU is better, with liquid cooling blocks, it's not like a battery that trickles up with heat, holds it, then dumps the heat. The goal is to get the heat to the water as fast as possible BEFORE the copper heats up to the point of causing the CPU to shut itself off.
So again, we only cool a cpu in a single 2 dimensional angle. Straight up. We don't utilize water to cool all of the copper surface area, that's why waterblock developers all make water blocks that cool nearly identical to each other.
2 dimensions? Fins on cold plates are 3 dimensions. CPUs are flat, we increase the surface area by going into the z axis with fins aka the 3rd dimension.
No one's thinking outside the box.
Beg to differ
You are taking advantage of just a tad more surface area. You are still only cooling copper on one surface of the waterblock. I have many design ideas for waterblocks that greatly differ from traditional.
On average, most cold plates are only being cooled by liquid on perhaps 30% of the potential cold plate surface area.
You can further drop temps with full metal blocks by simply cooling the remaining surface area.... with a fan.
Our cold plate fin surface area is 21.8 square inches. The rest of the plate is roughly 7.7in^2. So water is touching ~ 75% of the surface. 5.5in^2 touches the CPU.
So the only area that can be cooled with a fan is 0.99in^2.
No the extra copper doesn't help? Hmm, I'd beg to differ that statement while you are cooling maybe 1-2% more surface area than perhaps competition compared to your water block.
Definitely more than 1-2%
This is our upcoming threadripper block vs the top performing current block. You'll also notice our regular CPU block has MORE surface area than the competing threadripper block shown.
Metals store heat. Some can store more heat than others, some dissipate it faster than others. There's another thread here that would be of some interest while talking about an increased temperature gradient. Metals and liquids move heat faster with higher temps, your goal producing a waterblock is to lower the temps.
Not sure what you're advocating. Higher water temps?
I have an old school block. It's got a lot of copper. Doesn't even have fins, but I bet it weighs as much as yours (within 5%) and within 5c temps a decade plus old.
I'm not sure I understand. If your logic holds, then your old mega block with zero fins would perform better, right? I'm not sure what your liquid loop looks like now, but I'm guessing it has fins.
It seems strange you're arguing that surface area doesn't really matter.
Not assuming, knowing. There's a difference.
If there is space between the fins and the top plate, the water will take this path if it's greater than between the fins.
Water flowing through the fins will slow down from the friction against the fins.
The assumption then is there is a space between the fins and top plate. No waterblock has that space open. We definitely don't. Jet plates press down on that area.
Also, our fins are polished super smooth on the sides (impossible to see, of course) but the friction still wouldn't matter nearly enough to outweigh the massive increase in surface area from microfins.
I'd beg to differ and offer my lid-less PGA 2700x as proof this claim is totally false.
What are you basing this on? After all, you removed the copper IHS to achieve better performance. Less copper = better.
Pretty efficient design example, thank you.
So also strengthens the point of traditional waterblock cooling. Your only adding surface area inside a small water chamber. This does not utilize all of the copper effectively.
Then put Acrylic on top which is an insulator, while copper is never an insulator. It will absorb the heat and most evenly disperse this heat over most other metals.
With copper on copper blocks, some heat transfer and dissipation is lost from the bottom half to the top while there is a lack of a thermal interface material, perhaps some thermal padding to increase conductivity between the two halfs, then a fan on top like the one in the link there.
It's sad to solely rely on traditional waterblock designs. Most of them far from impressive in the sense that they are all basically the same.
Some use this fin count, some use just a small plate, some full copper..... but all of them actually utilize a small portion of the copper to water transfer.
So basically you're saying to add an air cooler on top of an all copper waterblock? Yes, that would improve surface area. And surface area matters. It sounds though like you don't believe water cooling surface area matters, only air cooling surface area?
