Here are some pictures of the revised (temporary) top for the nateman_doo chipset block, and some comparison shots of the inside vs. Mips.
This shot shows the bow in the temporary top. Although it straightened out pretty much when I screwed it into the copper bottom, it doesn't look like it's a perfect seal. nateman_doo said these turned out curved due to the thickness of the material from which they were cut. New and final top is already done and just awaiting drilling of the inlet/outlet hole locations.
Here are a couple of shots that compare the new top with the old beveled top. Same shape...just no beveled edge and change of hole locations....
....and here are a few pictures of the block with the new top on it. This block is huge and heavy.
Here are some shots comparing the size of the nateman_doo block with the Mips block. This block dwarfs the Mips block. nateman_doo on the left.....Mips on the right.
Mips on top of the nateman_doo block....
The nateman_doo block is larger in every direction. The other major difference in the looks of the blocks is that nateman_doo used black screws and Mips used silver screws. Which one looks better will depend on your build color and highlights.
These next 3 shots show the inside of the Mips block. Similar to the their VRM block....Mips uses a design that incorporates a thin flat copper plate....and all of the channel milling is done in the black top....not the copper. With the price of copper what it is....I'm sure this helps keep the price of the block down....but I would imagine that the way nateman does his block (channel milled into the copper bottom) would yield more temp transfer, with everything else the same.
Take a look at the water path in this Mips top. Wow. I think I've found out what made my loop so restrictive. Those are tiny little paths with VERY tight 180 degree turns in multiple places. Compare the size of the radius of the circle making those two small 180 turns in the corner with the radius of the inlet holes. They are about the same. So....imagine your loop having to do a 180 turn INSIDE of the threads of a G1/4 fitting. That is what is going on in there. TWICE. Even the inlet hole that comes in over the plastic jet plate side....is very very shallow. I don't want to knock this block....because my temps went from too high....to low 30's with this block. It performs beautifully. But the way it was designed has to be adding the most restriction by far to my whole loop. We'll see if flow rates increase when I swap the blocks out.
These next shots are of the nateman_doo block insides. It's pretty inside. You've probably already seen these from some of the videos he has posted along the way. You can see that the water channel is milled into the copper here....instead of the black top as Mips did it. I like this approach more.
The channel is deep and wide, and I wouldn't expect this block to have a lot of restriction. If I were to pick something to criticize about the design of the block...I'd say that the PLX chip closest to the bottom edge of the board is not going to get a lot of water circulation above it. In the picture below....its the nubby that sticks out on the farthest left edge of this picture. In contrast....Mips forced their flow around these corners using those tiny tiny little 180 turns. That keeps circulation over these PLX chip areas....but probably slows down flow rate overall....hurting other components. It's a trade-off....not sure which approach will work better overall. Might be a mixed bag. We'll see.
In this picture, the area in question is at the very top of the picture. You can see how it probably won't get much circulation. But then again....it probably doesn't need much. The PLX chips are not what gets real hot under here. Just having water in contact with the copper is going to transfer some heat....even with very little circulation.
This is a picture of the old Mips block on the board.
The Mips block is attached with five screws which screw directly into the block from the back of the board. Absorbed.....this picture is really for you since you'll have this block shortly.
I used paste TIM for the Mips block...and it looks like I had good contact with all 4chips.....
Skipping forward (forgot to shoot pics for awhile) to the installed nateman_doo block. The holes line up perfectly between cards in slot 1 and slot 3. They are spaced wider apart than the Mips holes were...but that allows nateman_doo to position them over the deepest parts of the channel.
You can see the hole positioning more clearly here....
I think the hole locations are just about perfect. You can see here that I can use a normal G1/4 fitting extender and it comes up cleanly between the cards without being close touching either of them. It might get a little tight if you have a backplate on the card in slot #3...but I think even that would work.
Even a big ol' Bitspower compression fitting sits cleanly between the two cards. But it is REAL close to touching...and if you go the compression fitting route, you probably will want to slip a little piece of clear plastic between the fitting and the cards to prevent it from accidentally shorting anything out.
So....I'm good with the placing of these holes...and I've let nateman know. Hopefully I have the final and permanent top for the block (no bowing) within a couple of days, and can test the temp performance and flow rate. Can't wait!