Random hard shutdowns followed by boot loop.

[Ethan_NLHW]

I just upgraded from a Z270 Chipset paired with an Intel Core i7 7700K to the X399 Chipset with an AMD Ryzen Threadripper 1950X. Followed the instructions to the letter. As the title suggests, I get random hard shutdowns followed by an endless boot loop. This only seems to happen when I am doing basic desktop functions. Gaming applications haven't given me any trouble thusfar. Below are the steps I have taken to correct the issue.

 

After the installation of the new board and CPU I performed a full OS re-install and wiped all of the drives.

Verified temperatures at the desktop with HWMonitor. Showing an identical readout on both packages, hovering around 34C.

Reseated both graphics cards.

Reseated the DIMMs.

Updated to the latest BIOS.

Installed latest chipset drivers direct from AMD.

Cleared the CMOS.

 

When it starts doing the boot loop, it will not stop. The only way I can get it to boot back into the OS is by throwing the main power switch on the PSU to off and hitting the power button to get the motherboard to completely power down. Sometimes not even that works. Other times it will shut off at random and stay off. When I go to turn it back on it will stay on, but the light on my SLI bridge will not illuminate and none of my monitors will get a signal. When the system is running, the debug LEDs on the motherboard read "AA". In the manual that code shows "The system has transitioned into ACPI mode. Interrupt controller is in APIC mode.", and I'm not entirely sure what that means.

 

If I forgot anything, feel free to ask for more info. System specs are in my signature. Thank you!

[ImNotThere]

thats a ram issue my friend

hard shutdowns frequently = psu

boot loops = ram

npt starting at all = mobo or cpu

[Ethan_NLHW]

Just now, ImNotThere said:

thats a ram issue my friend

hard shutdowns frequently = psu

boot loops = ram

npt starting at all = mobo or cpu

Could that be due to the DIMM slots on the board, or the modules themselves? They default to 2133MHz in the BIOS but are supported up to 3200MHz which is what I set them to with the A-XMP Profile #2 in the BIOS. Could it just be a memory compatibility issue due to Ryzen, and/or Threadripper?

[ImNotThere]

Just now, Ethan_JL said:

Could that be due to the DIMM slots on the board, or the modules themselves? They default to 2133MHz in the BIOS but are supported up to 3200MHz which is what I set them to with the A-XMP Profile #2 in the BIOS. Could it just be a memory compatibility issue due to Ryzen, and/or Threadripper?

ryzen ram compatibility is an old issue

could be dimms or dimm slots, check if the ram works in another system, if it doesnt then its ram if it does work, you have dead dimm slots

how many dimms are populated and which ones?

[Ethan_NLHW]

Just now, ImNotThere said:

ryzen ram compatibility is an old issue

could be dimms or dimm slots, check if the ram works in another system, if it doesnt then its ram if it does work, you have dead dimm slots

how many dimms are populated and which ones?

Four out of the eight on the board are populated. From left to right slots 1, 3, 6, and 8 have sticks in them. That's what the manual called for.

[ImNotThere]

1 minute ago, Ethan_JL said:

Four out of the eight on the board are populated. From left to right slots 1, 3, 6, and 8 have sticks in them. That's what the manual called for.

im pretty sure if one dimm/slot is broke it knocks the whole system, try one dimm at a time in slot 1 i guess

[Ethan_NLHW]

2 minutes ago, Ethan_JL said:

Four out of the eight on the board are populated. From left to right slots 1, 3, 6, and 8 have sticks in them. That's what the manual called for.

I should also specify that these DIMMs came out of the "old" Z270 board, and I never had any issues on that board.

[Ethan_NLHW]

2 minutes ago, ImNotThere said:

im pretty sure if one dimm/slot is broke it knocks the whole system, try one dimm at a time in slot 1 i guess

Alrighty. Would it be worthwhile doing a MemTest86 run? Also, could this be caused by improperly seated DIMMs? I've been typing this up on the system and haven't had it happen yet, and I just re-slotted the DIMMs before this (just got home from work). It started doing the boot loop thing instantly after I did that, but I wonder if that was something left over in the CMOS.

[ImNotThere]

7 minutes ago, Ethan_JL said:

Alrighty. Would it be worthwhile doing a MemTest86 run? Also, could this be caused by improperly seated DIMMs? I've been typing this up on the system and haven't had it happen yet, and I just re-slotted the DIMMs before this (just got home from work). It started doing the boot loop thing instantly after I did that, but I wonder if that was something left over in the CMOS.

possible but if its a boot loop id think its more of a dimm problem

[Ethan_NLHW]

15 hours ago, ImNotThere said:

possible but if its a boot loop id think its more of a dimm problem

Ran a Windows Memory Diagnostics test. Didn't find any issues. 

[DragonTamer1]

Boot loops can also point to issues with power delivery on the motherboard, usually the CPU. I doubt that a brand new board would have damaged VRM, but it is still a possibility.

 

A simple way to test this is to go into the BIOS and start disabling cores and clock the CPU and RAM at lower and lower speeds until it stabilizes.

[SpaceGhostC2C]

I would hold RAM and PSU suspects, in that order. I agree with @ImNotThere that the boot loop part points to RAM, but some other symptoms, like sometimes not starting, sometimes requiring a full un-powering (?) before starting on again makes me not leave the PSU out yet.

 

For PSU:

Which motherboard and PSU are you using?

 

For RAM:

Have you tried leaving the default 2133MHz? Does the problem go away if you do, or does it persist?

[Ethan_NLHW]

 

12 hours ago, SpaceGhostC2C said:

I would hold RAM and PSU suspects, in that order. I agree with @ImNotThere that the boot loop part points to RAM, but some other symptoms, like sometimes not starting, sometimes requiring a full un-powering (?) before starting on again makes me not leave the PSU out yet.

 

For PSU:

Which motherboard and PSU are you using?

 

For RAM:

Have you tried leaving the default 2133MHz? Does the problem go away if you do, or does it persist?

 

MSI X399 Gaming Pro Carbon

Corsair AX1200i

 

Also, yes it still persists at 2133MHz.

 

I did find someone in the Newegg reviews for that board that said they have the same exact issues that I'm having. That could point to it just being a BIOS issue with memory compatibility. 

 

It didn't shut down on me at all last night. When I woke up this morning and checked, it had shut down sometime last night and was unresponsive.

[SpaceGhostC2C]

1 minute ago, Ethan_JL said:

 

 

MSI X399 Gaming Pro Carbon

Corsair AX1200i

Stupid question, but: both 8-pin CPU connectors plugged correctly?

 

The rest of your post does sound like a motherboard issue anyway.

[Ethan_NLHW]

13 hours ago, DragonTamer1 said:

Boot loops can also point to issues with power delivery on the motherboard, usually the CPU. I doubt that a brand new board would have damaged VRM, but it is still a possibility.

 

A simple way to test this is to go into the BIOS and start disabling cores and clock the CPU and RAM at lower and lower speeds until it stabilizes.

This isn't something that is happening all the time. I could go hours without issue and all of the sudden it'll just shut down. It turns itself back on but is completely unresponsive. No monitor sync, and lights on the SLI bridge are off.

[Ethan_NLHW]

Just now, SpaceGhostC2C said:

Stupid question, but: both 8-pin CPU connectors plugged correctly?

 

The rest of your post does sound like a motherboard issue anyway.

Uh oh. Does it require two 8 pin power connectors? I only have one...

[SpaceGhostC2C]

2 hours ago, Ethan_JL said:

Uh oh. Does it require two 8 pin power connectors? I only have one...

Yes. If you only plugged one, then that's 99% sure the problem: insufficient power to the CPU. Your motherboard has two 8-pin connectors next to each other, and you PSU comes with the necessary cables to plug both, so unless you have lost some of the modular cables, you should be able to fix it  

(It will be labeled "EPS" or "CPU 4+4").

 

(EDIT: I dialed that back to 99 )

[Ethan_NLHW]

5 hours ago, SpaceGhostC2C said:

Yes. If you only plugged one, then that's 99% sure the problem: insufficient power to the CPU. Your motherboard has two 8-pin connectors next to each other, and you PSU comes with the necessary cables to plug both, so unless you have lost some of the modular cables, you should be able to fix it  

(It will be labeled "EPS" or "CPU 4+4").

 

(EDIT: I dialed that back to 99 )

Well I've got both CPU 4+4s hooked up and after two minutes on the desktop, it shut down again. Back to the drawing board on that.

[SpaceGhostC2C]

3 minutes ago, Ethan_JL said:

Well I've got both CPU 4+4s hooked up and after two minutes on the desktop, it shut down again. Back to the drawing board on that.

Damn, thought we got it

 

[Ethan_NLHW]

6 minutes ago, SpaceGhostC2C said:

Damn, thought we got it

 

Right before that I had a BSOD with error code "ATTEMPTED_WRITE_TO_READONLY_MEMORY". Too many answers on Google to figure that one out. I then shut it down and restarted and that's when I had the hard shut down again.

[Ethan_NLHW]

Bump

[xGGAx]

Have you done any testing with a different set of ram/ psu? We need to rule out that possibility.

If your ram and PSU work fine in some other rig, then your motherboard could be faulty.
I had a similar issue when I was in Tech support. We got an HP workstation that kept rebooting and stuff. First time it was a dead stick of ram. We removed it and it worked fine for a month. Then it started acting up again. All memtest, prime95 , etc. came out just fine.

We tested the psu and ram separately, they worked fine, but sometimes the pc would reboot itself or display a bluscreen. Long story short, the motherboard was killing the ram, one stick at a time. 
Im not saying that your motherboard is DOA, but it might be damaged somehow(again, assuming your ram and psu work fine).

[Ethan_NLHW]

18 minutes ago, xGGAx said:

Have you done any testing with a different set of ram/ psu? We need to rule out that possibility.

If your ram and PSU work fine in some other rig, then your motherboard could be faulty.
I had a similar issue when I was in Tech support. We got an HP workstation that kept rebooting and stuff. First time it was a dead stick of ram. We removed it and it worked fine for a month. Then it started acting up again. All memtest, prime95 , etc. came out just fine.

We tested the psu and ram separately, they worked fine, but sometimes the pc would reboot itself or display a bluscreen. Long story short, the motherboard was killing the ram, one stick at a time. 
Im not saying that your motherboard is DOA, but it might be damaged somehow(again, assuming your ram and psu work fine).

The RAM came out of a Z270 board that I had in this rig before. My PSU goes all the way back to when I was still rocking a Z97 chipset. Never had any issues until I put this board and CPU in. 

[Ethan_NLHW]

Well, I decided to try something. I booted up my rig yesterday afternoon and just left it at the login screen. Left it alone for about four hours and it never reset. Logged in to do some stuff, and it shut down. Decided to try I the same thing overnight, no reset. This leads me to think that this is something on the OS end. Thoughts?

[AMHDesign]

Hey, Just wanted to mention I have the EXACT same problem… And specs are almost the same too.

MSI x399

AMD Ryzen Threadripper 1950x

GTX 1080 ti

RAM: 4 x 16GB Corsair Vengeance LPX DDR4 2666MHz

PSU: EVGA 850 GQ

Windows 10 x64

Monitor 1: Wacom Cintiq 27QHD

Monitor 2: HP ??

 

I haven’t confirmed temp but fans are NZXT x52 which is suppose to use CAM (had to uninstalled CAM because extremely buggy and caused cursor lag etc…?).

 

Computer is used for content creation/ 3D rendering but both times (so far) the problem occurred when under low load (just on Photoshop or home screen).

 

Both monitors go black and cannot force shut down… have to flip the power switch at back.

First time I got a shock through my headphones.

 

Normally the LED lights display 40 after normal power-on or 30 after wake-up from sleep.

AA seems to appear after installing something and restarting or after problem occurs and am forced to cutting the power.