So
a user on the [H] forums did some tests of the LP53 in the Z270i STRIX motherboard with the fins perpenticular to the RAM, BUT the twist is he removed the rear I/O's VRM heatsink and the rear I/O panel, the test of removing that rear I/O VRM heatsink... dropped his core temps by 10c and eliminated the CPU's throttling at high temps!
He also noted that having the Rear I/O panel removed would allow CPU heat to escape more easily and keep temps down when the installed CPU Heatsink's fins are perpendicular to the RAM. EDIT: Then the user also seemingly had the same CPU coretemp drops from within the DAN A4 case:
Something to note: I think the orientation of the heatpipes themselves is important, in addition to fin direction. Once in the case and with the heatpipe bends of the LP53 at the bottom of the case, fins facing out the IO panel, I am noticing a further improvement in temperature. About 10c lower at idle vs open air and Max temp of Prime95 went down to 86c in the case, from 91c in open air! And that 86c was a singular spike, the average was more like 84c. So this matches conventional wisdom that fin orientation and heat pipe direction are both important, and wth the LP53 I think you can get away with the fin orientation facing the RAM so long as they can exhaust out in at least one direction, if the heatpipes are at the bottom of the case.
He is going to attempt to re-seat the LP53 with the fins parallel to the RAM, but he might have to remove the Z270i's Wifi card enclosure in addition to the rear I/O heatsink, as he believes doing so will allow the LP53's heatpipes to fit by the Rear I/O. I havent recieved mine yet... BUT!
Seeing this user gain a temp decrease from removing the Rear I/O VRM and the rear I/O panel... from inside the Dan A4 case,
this was a lightbulb moment for me. I had never thought to do this, but apparently people in the DAN A4 community have been doing this and turning out positive results! So here's my thought, I'm going to test this out on my current NHL9i, as it sits now, its installed with the fins parallel to the RAM...
BUT I'll retest coretemps with the addition of removing the Z270i STRIX's TOP VRM heatsink (by the 8-pin CPU power)
as it's blocking one end of the NHL9i fins... and seeing how it's removal affects CPU core temps at idle and at load (and praying for a significant temp drop) at 4.8GHz OC, stock vcore on my 7700k.
All of these tests will be conducted from within the S4 mini. And if anyone has any questions about the safety and if the 8-phase power MOSFETS/VRMs will be able to run at or below peak/warning operating temps, I found some research on reddit
here in a post where someone was asking about removing the VRM heatsinks on their Z270i STRIX board. According to that thread on reddit, the VRMs should be fine when running under 120c, and according to the mobo intro part of
[H]ardOCP's review of the Z270i STRIX, the VRMs were running at a cool 100-105c with the VRM heatsinks on top, so the Z270i should be fine if I take them off, as long as I don't OC.
Also, once my LP53 arrives, I'll be testing it out with as many orientation configurations as possible, trying to find the best one that can outperform the NHL9i... both coolers using the NF-A9x14 with the same fan curve I've used this whole time during my temp tests.
