High-Load DRAM & Q-Code 8

[pankomputerek]

I've seen many conflicting sources as to the q-code 8 troubleshooting some say it's cpu and some ram issue.

I keep crashing constantly when the PC is put under high stress for example gaming (easy replicate in minutes in Rust for example). The screens will go static and the system will shutdown with Q-Code 8 and DRAM led flashing. I'm using Gskill F4-3000C15Q-32GVRB (4x sticks 8GB, at 2133mhz). Currently testing 0902 bios however issue is persistent across all bios versions.

Tried manually changing dram voltage at stock speeds and the timings picked up by bios are what they should be.
Memory Test comes out with no errors from Windows.

I know few people had luck with changing soc voltage, what should I keep it at? I'm using 1800x watercooled.

EDIT: It would appear that the GPU that I have is crashing the system after extended load. I am yet to resolve the issue. Tried drivers and different power supplies with no luck.

[Kobura]

Is your 1800X overclocked? Q-Code 8 with DRAM LED flashing is what i get when my PC crashes from too low vCore on overclocks.

[pankomputerek]

Kobura wrote:
Is your 1800X overclocked? Q-Code 8 with DRAM LED flashing is what i get when my PC crashes from too low vCore on overclocks.

I run it on stocks speeds, AI Suite is set at 100x, 3621.6mhz.
I don't care much for overclocks if system is not stable at first, AI Suite reports CPU core voltage of ~0.763v. Should I change it in bios?

[Kobura]

pankomputerek wrote:
I run it on stocks speeds, AI Suite is set at 100x, 3621.6mhz.
I don't care much for overclocks if system is not stable at first, AI Suite reports CPU core voltage of ~0.763v. Should I change it in bios?

If you're still on 0902 you can try offset mode with a small value like 0.0125V on the vCore. But it actually should run fine with auto on stock settings.

Maybe try running memtest86 on your kit, could be a faulty RAM module. Though they usually crash with a BSOD instead of going the Q-Code 8 route... still worth a try. Or try the same stress test with only two modules in 2nd and 4th RAM slot counting from the CPU.

Also for voltages:
VDDSOC: try up to 1.15V, but it's preferred if you can stay below 1.0V. Definitely don't go over 1.2V.
RAM: you can try up to 1.4V if you're on Samsung B-die. They will easily take that on ambient for daily usage. I would try up to 1.38 on yours, since it's highly likely they're not Samsung B-die.

[pankomputerek]

Kobura wrote:
If you're still on 0902 you can try offset mode with a small value like 0.0125V on the vCore. But it actually should run fine with auto on stock settings.

Maybe try running memtest86 on your kit, could be a faulty RAM module. Though they usually crash with a BSOD instead of going the Q-Code 8 route... still worth a try. Or try the same stress test with only two modules in 2nd and 4th RAM slot counting from the CPU.

Also for voltages:
VDDSOC: try up to 1.15V, but it's preferred if you can stay below 1.0V. Definitely don't go over 1.2V.
RAM: you can try up to 1.4V if you're on Samsung B-die. They will easily take that on ambient for daily usage. I would try up to 1.38 on yours, since it's highly likely they're not Samsung B-die.

Thank you for prompt replies! I am currently experimenting with the settings you recommended.

After some tinkering around it looks like the issue *might* be related to GPU drivers. It seems to happen after about 10mins of furmark. Will change drivers and report back.

EDIT: 2nd test on new 17.3.1 (up from 16.x - After 20 minutes of full load system packs in.

EDIT 2: Video of the issue along with new thread as it might not be MoBo related: https://rog.asus.com/forum/showthread.php?91694-DUAL-RX480-O8G-crashing-during-gaming-stress-test&p=...

EDIT 3: I have tested another power supply same issue is recurring.