[SOLVED] Asus Z370 system crashes -- won't post again until I change the RAM config

[cfarley137]

My kid's gaming computer was working fine for a few months, but recently started crashing. It basically just shuts down while he's playing GPU-intensive games. The power LED remains on after a crash, but the fans and everything is off.

After a crash, the computer won't post at all until I remove one of the two DDR4 modules. Then the computer will boot again. I can switch the slots (moving one module from slot 3 to slot 4 for example). There's rarely anything interesting in the event viewer. Except a few blue screen dump files have been generated in "C:\Windows\Minidump". I processed the 3 or 4 I have found and get references to VIDEO_TDR_FAILURE

I've recently re-installed Windows 10 and the issue still occurs, so I'm now 100% sure this is a hardware issue. (Which is what I suspected anyway.)

I've tried making the computer crash with stuff like Prime95, FuMark, a Linux-based RAM thrashing utility called Stressapptest, etc. I have not been able to force a crash. But let that kid spend a few hours playing Fortnite and it will eventually crash.

Here's a list of components:

MOBO: Asus Z370-PLUS TUF Gaming
PSU: CORSAIR RM750x
CPU: Intel Core i5-8400 2.8 GHz 6-Core
RAM: 2 x 8GB sticks -- G.Skill F4-3000C15D-16GTZR
GPU: ASUS ROG Radeon RX 560 STRIX-RX560-O4G-GAMING 4GB
SSD: WD Black 512GB Performance SSD - 8 Gb/s M.2 2280 PCIe NVMe Solid State Drive

The RAM is not on the Motherboard's QVL list, although G.Skill says the RAM is indeed compatible with the motherboard. I've run some hardware monitoring utilities (such as GPU-Z) and have not found any weird voltages or thermal issues.

I guess now I need to start swapping components to find out what's broken. I'm not sure how to do this strategically. What should be my next step?

 

[MeanMachine41]

Hi cfarley137

I would run a test on RAM using Memtest86 and booting from the USB. run the test 2 or 3 times on all modules and see if there are errors. If there are any errors then RMA the full kit. If there are nill errors then you may have stability issues under load. I use AIDA64 for testing both CPU and RAM when system is under load. Try it it's free for a month. Many manufacturers specify compatibility which is not always the case. Best are kits on the MB QVL that are tested by the MB manufacturer.

I have G.Skill F4-3000C14D-16GTZR (2x8) and had to tweak DRAM timing in Bios for stability. Profiles did not work as 3000MHz is OC RAM.

 

[cfarley137]

Hi cfarley137

I would run a test on RAM using Memtest86 and booting from the USB. run the test 2 or 3 times on all modules and see if there are errors. If there are any errors then RMA the full kit. If there are nill errors then you may have stability issues under load. I use AIDA64 for testing both CPU and RAM when system is under load. Try it it's free for a month. Many manufacturers specify compatibility which is not always the case. Best are kits on the MB QVL that are tested by the MB manufacturer.

I have G.Skill F4-3000C14D-16GTZR (2x8) and had to tweak DRAM timing in Bios for stability. Profiles did not work as 3000MHz is OC RAM.

Thanks for the suggestions. I've run Memtest86 several times and it has not detected a problem. Also Stressapptest and the Windows Memory Diagnostics. I've yet to see an error.

I appreciate the suggestion to do a trial of AIDA64. Up until now I've been running stuff like GPU-Z, logging to a file, and then using a tool to look for unusual voltages or temperatures just prior to a crash. It looks like AIDA is a bit more hands-off.

 

[MeanMachine41]

I appreciate the suggestion to do a trial of AIDA64. Up until now I've been running stuff like GPU-Z, logging to a file, and then using a tool to look for unusual voltages or temperatures just prior to a crash. It looks like AIDA is a bit more hands-off.

Aida64 will stress your system far more than any game and provide diagnostic information that can determine which piece of hardware is failing. As your 100% sure, run the stress tester for 10mins with check boxes for CPU, FPU and Cache. If you pass that and all is OK then do the same for RAM, GPU and Disk.
You would be looking for any temperature and Rail Voltage anomalies under load.
During testing I run Aida64 along side HWinfo64. You can post screen shots of the information for analysis here

 

[mehdi2m]

Hi cfarley137, do you fix the problem?
I have the same problem with my Asus z370 TUF pro gaming.