ROG VI APEX detects only 2/4 Ram Modules correctly?

[j_battermann]

Hello there,

I got a somewhat strange problem with my new ROG Rampage VI Apex board in combination with a 4-Dimm (4x16gb 3000mhz) Corsair Vengeance DDR4 3000mhz kit: the board has its latest bios applied (September 15th 2017) and all 4 ddr4 modules are installed on the board, however the board/bios and therefore also OS only detect 2 of the modules correctly and consequently only 32 out of the 64 expected GB.

Detecting correctly as in - inside the bios the SPD information displays the full information for two of the modules only, for two other modules it displays the information only partially (see attached screenshots), albeit some of their information is there.


Now I tried shuffling the modules in the slots but no change - 2 out of 4 are fully detected.. two are not. I also re-tried all 4 modules in the original system I pulled them from .. and all 4 work perfectly fine there. I also tried loading the bios defaults - no change.. the ROG Rampage VI Apex does not properly recognize all 4 of them.


Does anyone know what's going on here / how I could get the board to recognize all 4 of the modules?


Thanks,
-Jörg

[j_battermann]

What's maybe also relevant here: when I remove two modules out of the system and just run it with the remaining two modules (and no matter which of the two.. I also tried multiple combos), it only works if I stick them in the DIMM_A1 and DIMM_B1 slots. The manual says if two modules are used, they are supposed to go into DIMM_C1 and DIMM_D1.. but if I do that, I get a Post Code of 53.

I can freely shuffle/mix the modules around and as long as I stick two into _A1 and _B1.. it does work.. but never in _C1 and _D1.

I checked and checked whether there's something in the slots that might 'block' or something the modules from properly getting contact.. but nothing is visible. Took out the vacuum and carefully went over the slots.. but no change.

So.. yeah.. I am confused and worried that it might be the board itself.. physically. Unless someone says that there's some magic jumper and or button or something I need to press/hold or has any other idea.. 😕

-J

[jab383]

I guess you're using a SkylakeE CPU that is supposed to see all 4 DIMM slots. The 7740X is a 2-channel CPU that will only see slots C and D, but slots A and B are the ones working right.

Which of the DIMM LEDs are lit? See page 1-20 of the user guide.

Slot pairs AB and CD have separate power regulators and voltage settings in BIOS. Check Extreme and Tweaker's Paradise tabs of BIOS to see that both slot pairs are set the same. In the DRAM settings tab, there settings for turning off DIMMS slot by slot. Check that they are all enabled.

[j_battermann]

The CPU is a 7960x.. so yep - it should support all 4 dimm slots. I checked the BIOS and all voltage etc settings in Tweakers Paradise are set to Auto. Led wise on the mobo the AB ones are lit when only those two slots are filled, I'll need to check what they show with all 4 slots filled tomorrow (after disabling XMP again... with XMP enabled and all 4 modules the Post code led displays shows a '6' and... erm.. an inverted/mirrored '6'?!? and stops there.. and since it's past midnight here, I'll do it tomorrow).

[Praz]

Hello

Check the CPU socket for bent or damaged pins. Also be sure the cooler/water block is properly tightened when reinstalled.

[xarot]

Just one thing I have noticed from the problem you are seeing: I was having serious similar problems with my old Rampage V Extreme where the C1 slot was not detected. As after hours I figured out, it was actually the B1 slot that needed some fiddling to get the C1 working too. I guess you took a flashlight and checked all the DIMM slot pins that none of them are misaligned? Also often screwing the cooler or water block too tight will cause issues if the board bends even a little. I first saw this issue in Asus P4C800-E Deluxe Pentium4 motherboard.

Try the C1 slot alone, I am unsure if the Apex will post with no DIMM in A1 but try it with all the sticks to narrow the problem down.

[j_battermann]

Good evening everybody,

thanks for the help and suggestions! You had me worried regarding the potentially bent/damaged cpu socket pins - so I removed the cooler again, removed the CPU and very closely checked.. but nope - all pins are fine. Re-installed the CPU and cooler.. this time taking particular care regarding tightening the cooler 'enough' but not too much.. checked whether there is any visible stress / bending or anything on the board.. but no. I also double checked the dimm slots themselves.. but no - all looks good.. re-checked.. and still.. everything looks perfect visually. So I tried again - but no change.. same problems as originally mentioned.

I tried one dimm in C1 only (all 4 of them) - no change (as in.. POST Code 53 and that's it).


I also had the idea this morning to get another set of 4 dimm modules to check whether it has something to do with mine and get a 4x8gb pack of 3200mhz (unfortunately also only corsair ones were available for same-day delivery here.. but it's a different series / product number and also overall size.. so I figured it might be worth a try just to check whether or not it is somehow related to my set of dimms) .. buuut nope - exactly the same behaviour with those different set of dimms. C1 & D1 dimm slots just refuse to work as expected.. well.. or at all.


Regarding the LEDs as mentioned.. the ones shown in the userguide at 1-20: whenever I use the C1/D1 slots.. none of them light up. However there's also a led strip directly underneath the Post-Code LED display.. next/close to the LN2 jumper and that one does slowly blink when C1/D1 slots are used.. and Post Code 53 is shown.. but that's it.

When using the A1/B1 slots (only)... that LED shines constantly.


Anything else I might/could try? 😕


Thanks,
-Jörg

[j_battermann]

Ok now this is getting weird - I got a second APEX board (Asus doesn't do direct replacements here in Germany / with this board).. built the system up & again - exactly the same problem.

However, what I just noticed when all 4 modules are in the system (and the bios/spd tool in there reports only partial information) and it boots into windows (with only 2x16gb / 32gb 'visible' to the bios & windows), HWiNFO64 does also display all 4 dimms, including their -full- SPD etc information (see below).


Could this be caused by..

• the bios maybe not supporting the modules (even though both sets that I've tried are on the board's QVL list for Skylake-X CPUs?)
  
• some sort of voltage setting for the modules in the bios / 'Auto' not being sufficient / correct
  
• my power supply (it's a 550w be quiet dark power) which 'should' be enough given the TDP of the components in the system.. as in the CPU being.. what.. 165w.. the GTX 1070 having a 150w TDP.. and the rest is 'just' 2 Nvme SSDs, 2 2.5" ssds and an AiO cooler (also bequiet).. but I am wondering whether there might be a 'spike' in power-draw when the system is turned on that goes 'over' the power supplies capacity..?
  
• The CPU itself? It's a delidded one (not done by me but by caseking.de / the place de8auer works at) .. so what I am wondering is whether it might be possible that some hardware on top of the cpu / under the ihs get slightly damaged or shortened or whatnot.
  

Just thinking out loud here.. but maybe someone has some knowledge whether one or the other does actually make sense and/or what to do/try next...

[Dancop_HWBOT]

j.battermann wrote:
Ok now this is getting weird - I got a second APEX board (Asus doesn't do direct replacements here in Germany / with this board).. built the system up & again - exactly the same problem.

However, what I just noticed when all 4 modules are in the system (and the bios/spd tool in there reports only partial information) and it boots into windows (with only 2x16gb / 32gb 'visible' to the bios & windows), HWiNFO64 does also display all 4 dimms, including their -full- SPD etc information (see below).


Could this be caused by..

• the bios maybe not supporting the modules (even though both sets that I've tried are on the board's QVL list for Skylake-X CPUs?)
  
• some sort of voltage setting for the modules in the bios / 'Auto' not being sufficient / correct
  
• my power supply (it's a 550w be quiet dark power) which 'should' be enough given the TDP of the components in the system.. as in the CPU being.. what.. 165w.. the GTX 1070 having a 150w TDP.. and the rest is 'just' 2 Nvme SSDs, 2 2.5" ssds and an AiO cooler (also bequiet).. but I am wondering whether there might be a 'spike' in power-draw when the system is turned on that goes 'over' the power supplies capacity..?
  
• The CPU itself? It's a delidded one (not done by me but by caseking.de / the place de8auer works at) .. so what I am wondering is whether it might be possible that some hardware on top of the cpu / under the ihs get slightly damaged or shortened or whatnot.
  

Just thinking out loud here.. but maybe someone has some knowledge whether one or the other does actually make sense and/or what to do/try next...

I've had the same issue with 3 chips now...and no, it's neither the board, nor the dimm!
Let me guess, you deliddded your chip, no?
If yes, there's one resistor, quite near to the die, missing then...

[j_battermann]

Dancop_HWBOT wrote:
I've had the same issue with 3 chips now...and no, it's neither the board, nor the dimm!
Let me guess, you deliddded your chip, no?
If yes, there's one resistor, quite near to the die, missing then...

Yep - it's a delidded one. I didn't do it myself but I was suspecting that something might have gone wrong. They said they tested it prior to sending it out, but due to the nature of this problem, maybe they didn't test with 4 dimms etc.

Oh my.. I'll have to send it back then. Would you by any chance happen to have made a picture of the (location of the) resistor you've mentioned so I can let the caseking.de folks know where to look specifically? Or maybe point out the one as seen on the delidded 7980xe (mine's a 7960x.. but I assume they look similiar)?