Hey everyone,
I am at a complete loss and at the end of my logic (and sanity):
Once upon a time,..
I got myself a 14900KS and a competitors high end enthusiast board for my new direct die build. After some 4 weeks the whole thing rapidly descended into freezes, BSOD bootloops and then a terminal error code 55 dead end.
Since the whole microcode thing was going on back then and the symptoms fitted, I grudgingly "accepted my fate" and decided to....try again. So I got basically everything new: A brand new 14900KS, an Apex Encore and new QVL compliant G.Skill RAMS (Dual 48GB 6400 kit).
At this point I would like to point out, that the first rig turned out to still boot normally on one DIMM in one slot (A2).....everything else -> Error Code 55.
Imagine my shock, when after putting everything together and literally standing in front of a whole new rig......it goes to 55.
Now to cut a long story short because, believe me, I spent a week checking pretty much every option that was put online within the last 5 years, I am absolutely clueless, and I have been building rigs for 30 years now.
1. I updated the BIOS to latest via USB Flash
2. Both DIMMs work individually in DIMM slot 1.
3. No chance of getting to Post with 2 populated DIMM slots. DIMM 2 gets power however.
4. None of them works individually in DIMM slot 2, at all.
....at this point: All tests and retries where done with two RAM kits just to make sure. Both QVL listed.
5. I removed and remounted the cooler (EKWB DD Quantum) 5 times. I tightened, loosened for hours, no change apart from not working at all if the cooler is too loose.
6. When I insert only one DIMM in slot A everything works fine, the DIMM gets recognized perfectly and the system booted into both XMP settings straight off the bat without any tuning. Windows launches and works as intended. Also, CPU, DIMM, Board and GPU show (expectedly) remarkably low temps (1000D... full of fans and radiators and individual component loops and whatnot)
7. Absolutely 100% no bent pins on the socket. Not one that even remotely makes a doubtful impression. Zero. There is no visible physical damage or even irregularity o be found....anywhere.
8. Changed to the ThermalGrizzly Mycro Direct Die cooler because of getting really really paranoid about mounting pressure. No change. 3 attempts.
9. Reset, Unplug, Clear CMOS, remove battery.... the works. Nothing.
10. Everything reads out as intended in AIDA, CPU-Z, HWMonitor etc., with one DIMM inserted.
Since I have literally swapped out every component now except for the PSU, I am close to losing it.
How high is the probability of either killing two CPUs during delidding in the exact same way (in terms of damaging the Mem Controller??? Both delidded CPUs show no sign of any physical damage (another hour with the magnifying glass) and both booted into Windows etc. when only one DIMM was being used (on the Apex Encore as well as the previous board). No sheared off SMDs, no visible damage to two mirror shiny dies. The foam "washer" needed to insulate the SMDs from the liquid metal was renewed each new attempt. No spillage, no traces of any liquid metal where it shouldn't be. No board creaks, no slips during assembly (being paranoid by now as I said).
How high is the probability, that a brand new board miraculously has a (pretty much identical) DIMM lane issue, out of the roughly 30 other it could have that I could think of?
If anyone has any idea it would be galactically appreciated. I simply have no remaining logical theory that does not include lottery like (im)probabilities left.
Naturally, if I missed out on any relevant info that could be related to this mess, please let me know, I will try to reply as detailed as possible.
Thank you in advance!
K.
