Good morning/day/evening EVGA community,
 
Just recently I had experienced a GPU failure (no image output on multiple systems) which is to be RMA‘d this week.
 
Now, after the GPU had died I am having trouble with my system RAM(Ballistix Elite 16GB (8GBx2) DDR4-4000 UDIMM) or(?) the motherboard (EVGA Z390 Dark) running the XMP profile.
Whenever I try to run 2 sticks with XMP enabled, which worked like a charm ever since I bought all the parts and put the system together, I am getting no video output and it shows a few post codes.
 
Sometimes but not always it even leads to (that red LED):
CATERR - Catastrophic Error on the processor a. RED: Processor error has occurred.
 
Now the system has a EVGA GTX 1060 in it and it and everything is stable with 1 stick but not with 2 sticks (XMP = no boot).
 
What could be the reason for it? RAM bad? Motherboard bad? CPU degradation(runs fine at stock and 1.250 vCore / -0.050 Core Voltage Offset)?
 
Thanks in advance for any replies!
 
EDIT: Both RAM sticks passed full memtest86 run when being used individually.

Cpu imc degration would be my guess.

Sajin
Cpu imc degration would be my guess.

 
Thanks for the quick response Sajin.

It happened exactly the same time the GPU died. Could it be correlated by any chance?
This is how it happened:
PC cold start (first time of the day), 5-10 minutes of browsing(YouTube, text pages etc), screen freeze, restart, no output, GPU dead, GPU replaced, no output, 1 stick removed, PC starts with output.
 
Is there a way to make/be sure it is the IMC by any chance? I'm not even sure what's the gurantee situation on a 9900KS. 

No problem. No way to really check unless you don’t mind using massive amounts of voltage (which could also cause more degration issues) to see if you can get it going again because the imc may need massive amounts of voltage to make it work properly now. 1.3v vsa/vccio is the most voltage you should use.

Sajin
No problem. No way to really check unless you don’t mind using massive amounts of voltage (which could also cause more degration issues) to see if you can get it going again because the imc may need massive amounts of voltage to make it work properly now. 1.3v vsa/vccio is the most voltage you should use.

Yes, wouldn't want to go higher than that. Been thinking to get it replaced (CPU) since it is within the 1 year period.
 
Just checked on one of my first threads here on EVGA with out-of-the-box readings. IO and SA were 1.32~ and I think it might have been running like that for a while.

Let us know how the cpu replacement goes.

Just read the vCore from the debug LED (on the motherboard) and it displayed 1.428 voltage when doing CPU-Z stress test. The debug LED should be the most accurate, right? Stopped it right away of course. Everything at auto (nothing changed in BIOS).
Then I just set the IO SA to 1.3v for a boot and it made the WiFi chip/function disappear completely. Reset CMOS and it is back to life at all stock/auto.
Everything still leads to CPU?

Yes, if both sticks work fine at default non xmp values when both are plugged into the board.

Yes, both in - working without XMP.
I guess XMP is not that safe to use after all (in some cases) and lesson was learned here to do everything manually so IO SA doesn't get shot up that high.
 
Been just reading:
 
https://www.intel.com/content/www/us/en/support/articles/000054907/processors/intel-core-processors.html
 

INTEL
Note 
Altering the frequency and/or voltage outside of Intel specifications may void the processor warranty. Examples: Overclocking and enabling Intel® XMP, which is a type of memory overclocking, and using it beyond the given specifications may void the processor warranty. If an overclocked processor was enrolled in the Performance Tuning Protection Plan (PTPP), the protection plan will cover processor replacement only, not memory replacement

 

 

The i9 9900KS has more power headroom than any other CPU for that socket. Also, in my experience, the 3 9900KS I tested were way more capable of powering the 4000MHz+ RAM than any other chip I had. The two i9 9900K's I attempted to run my 4266MHz CL17 RAM on, didn't even boot... I even tried pushing the voltage to 1.55v and blast a fan directly onto the RAM, had more luck with a i7 8086K...

I would suggest disconnecting any other hard drives except for the drive running the OS. I've had HDD's cause boot failures in the past and allow a successful boot intermittently but it could be a failing drive.
 
I would consider checking the PSU, you moved from a RTX 2080Ti over to a GTX 1060 which takes considerably less power to run, you can monitor that with software like PC Wizard to see how your voltages look when you stress your hardware. 

I've had a similar problem like that on my i9 9900KS when I had it... it even happens on my i9 10900K. Sometimes a windows update will mess with my overclock and I have to figure out the new voltage requirements. I would suggest setting the XMP profile and after the profile is enabled, manually set the voltage up .5 from what the XMP profile uses (Example: XMP uses 1.35v so set to 1.4v.) That is what fixed my problems. Also, be aware... that's voltage range where you need to watch your temps, make sure those RAM sticks are getting some decent air flow.
 
**SIDE NOTE**
I sold my i9 9900KS cause it was worth way more than what I paid for it. I would consider that as an option to move over to Z490. I posted mine on eBay for $775 and it sold in less than 2 hours.