Dotatech "accidentally" sold a 5950x for RM2.2k few days ago on Shopee. Whoever copped it, good freaking steal.

Haiz, I ran into problem again on my system just under a year.

My PC shut off and restart randomly during normal operation.
The blackout is fast one, just a second before it alive, but it won't boot with Qcode 00 (not used code!!!) and QLED stuck at DRAM, and system not response to power button long press either. Have to kill the AC switch, then it will boot normally.
It will even reboot sometimes in BIOS setup page.

Try swap my RAM, still happen.
PBO disable, not fix.
Even run at optimised default also reboot.

I suspect now is due to either CPU or mobo, with highest suspicious on CPU.

Can you guys suggest a sensible way to further test out whether is CPU or mobo? I'm frustrating now as i just recover from a nightmare failure on my previous system before i switch to AMD platform, and the seemingly HW issue haunt me again.

I just want to avoid buy something new just to isolate the factor, but i won't know exactly the issue if i didn't do so, contradict, arggg

update BIOS

Where's your logic?

The problem isn't occur because of BIOS change previously. It worked fine for last 8 months until recently. So what to do with BIOS?

Hi What is the new price for Ryzen 9 5950X?

Restore ur windows to previous month . Just double check this few days what have you done

No, do nothing.
If youbread my post, it even reboot in BIOS setup page, nothing to do with OS.

That reboot isn't BSOD, it just shut itself off briefly, wake up and stuck, looks to me a HW failure, not caused by SW

If you've ascertain it's nothing to do within the OS, then it's hardware.

You've ruled out RAM, fine. What about the RAM slots? Since you said you swapped RAM and it still happens, it COULD be due to faulty RAM slots. Yes it sounds very unlikely and silly why something you don't touch would fail, but it does happen.

The other thing to check is PSU, yes as unlikely as that sounds - PSUs can cause random reboots. I've certainly ran into a few random reboots due to a failing PSU.

Of course worst case scenario is the CPU, which hopefully isn't the case.

coolkwc
This
if it happens during high system stress. An aging PSU can get overwhelmed by transients pretty easily.

The problem can be consistenty reproduced when doing system memory stress in AIDA64, which lead me to believe is RAM issue, but swap with different pair of RAM didn't solve it. Your point is valid, i'm currently stress it under single channel and let see what's the outcome. My RAM only can run in the current slot if run in pairs, run in another dual slot won't boot due to the design.

As for the PSU, even if it failed, the system should have boot normally when power resume rather than stuck. The thing is, even the long press power button shutdown is not work when it stuck, i have no idea why and i believe this is an important crue as well. The system simply locked unless AC killed.

This post has been edited by coolkwc: Apr 20 2022, 05:53 PM

My another question would be, if let say i able to proof the RAM slot/mobo or CPU is the root cause, how do i proof to the seller if its random? It would be much easier if it just dead rather than random.

Does it random reboot if you do any other benchmarks?

I mentioned RAM slots because back on my Intel systems (Haswell based, Z87), one of my RAM slots became faulty out of no where eventhough the RAM I was running weren't OC'd nor running out of spec. The system still works and will boot normally as long as I NEVER cold boot it via switching off the PSU switch. Otherwise the system will go into random reboots or just simply fail to POST. Switching the RAM sticks to a different slot fixed that problem.

PSU is a little more complicated. The ATX 24 pin has a pin that reports the power sequence between motherboard and PSU, sometimes this can bug out due to abnormalities in the PSU and/or the mobo which causes the system to lock up because neither the mobo nor PSU can "communicate" anymore.

This post has been edited by SSJBen: Apr 20 2022, 07:08 PM

After some re-thought, I believe the problem embedded since long time ago.

Normally system crash/reboot is just suddenly black out and re-POST, but seldom have the symptom that I experienced now, which is shut off (can see all the light turn OFF, and my sound card and PSU relay release click sound) for about 1s, then reboot, without POST/beep sound, and just stuck with Q-CODE 00 and QLED pointed at DRAM.

This is how it happen, and it actually happened since Oct last year but just that i don't know what causing it.

Oct/21 - I start Ravencoin mining, lasted for 1 week. Everything works fine until after a week, my PC shut off and reboot and stuck, this is 1st time it appear. I thought is due to GC OC, so just limit the GC power to 85%, and the crash didn't happen anymore. Mining halted eventually few days after this event, as i don't want to spoil my rig unneccesary.
Nov-Dec/21 - I remembered i have asked someone in Nvidia thread anyone experience crash to desktop without warning during FH5 gameplay. I tried default clock for my GC and put my AMD CPU PBO to auto, but crash still happen. There are few times the system shut off and stuck after reboot.

But up until this week, the shut-off -> reboot stuck event won't happen in desktop idling. It only happen since Monday. It will just occur in idling, movies, just randomly. Stress test using AIDA64 is able to crash it.

Just now i tried remove 1 stick and swap between A1 and A2, crash still happen in both slots. So now i can definitely rule out RAM and RAM slot.

I think now left 3 potential root cause, either CPU, MOBO or PSU. I still have a spare 600W PSU to try out later, but if end up not the issue, i will left with 2 items which is very difficult to proof.

After read the issues you have described, maybe you should test with your spare PSU first. The reason i say so is because i have experienced some similar symptoms several months ago, end up it was my Corsair AX1200i issue. I read from other forums/groups many also having issue with the old AXi series PSU after they changed to high end GPU.

Anyone here has encounter this issue ever?

Even a hardware reset button isn't work!!! When i press the reset, the system just change from stuck 00 code to AA, and a HW reset just won't reset the system, and must kill the AC.

No, i think is MOBO issue here. No reason the HW reset button won't be able to reset, it just drive me crazy.

Would it kill you to just test with your spare PSU and eliminate 1 factor entirely?

nrwSSJBenllk

I think i found the problem already, which hint by the Asus sound card.
The sound card would required external supply from Molex 4 pins and will prompt if is not.

So somehow after the reboot the prompt appear, i'm like impossible because that 4 pin Molex is very tight.
This trigger me potentially is a PSU issue.

So i fire up the stress test again, and monitor the voltage level from both Corsair link and HWinfo64. The Corsair link reported 12V and 5V is extremely stable even under 600W+, what surprise me is the mobo reading in HWinfo64 shown only 11.56V and 4.65V from both rails.
Immediately i open the casing and measure the 4 pin molex at sound card using digital multimeter, the 5V is measured around 1.8V and 12V rail at 11.78V. Once i slightly pull and push the connector, the 5V rail now measure around 4.96V.

Without changing to spare PSU, i guess it could be due to either loose or oxidised contact on both the mobo and PSU end. So what i did was unplug all the cable connectors on both end and replug it for 3-4 times try to let the contact friction scratch away oxidation layer if any.
BINGO, after done all the multiple unplug/replug work, the 12V and 5V reported by mobo during stress test now become 11.98V and 4.96V, and no more shutdown/stuck reboot issue almost 2 hours now.

I think the problem is the contact resistance increase due to pin oxidation, at both PSU and mobo/GC end, especially PSU end which i don't really unplug since i built the rig few years ago with Intel platform, and just change to AMD platform last years without touch the PSU end connection, which in turn causing excessive voltage drop when high current pass-through. This is why something you don't touch would fail like what SSJBen mentioned.

Lesson learnt, occasionally touch it even is unnecessary.

This post has been edited by coolkwc: Apr 21 2022, 01:17 AM

Glad it all worked out for you.

Sometimes the most unlikely part to fail is exactly the part that fails.

So I finally jumped into SFF, with a NR200P for my new build - my son's PC one day just refused to boot no matter what I throw at it (old FX6300), thinking its not worth salvaging, so I decided to give my 2600 to him, and make a new one for myself. Yes it's nothing high end, some stuff I got at sale + local pc shops

Ryzen 5 5600
Thermalright AXP90-X47 Full copper
Asrock A520M ITX-AC
Silicon Power 8GB x2 DDR4
Superflower Silver Eco 500W
silly R5 240 gpu (temporary..)
in NR200P with 4 Sickleflows
Latest BIOS, chipset drivers
Balanced power plan

So immediately after booting and updating drivers etc.. I noticed the idle is on the bit high side (50C+, 40++ only during cooler nights/morning), and it vey quickly shot up to the 95C limit even with simple CPU-Z bench/stress test. Even running SignalRGB at the background keeps the idle temps at 60C+, hence I've deleted it and will switch to a hardware ARGB controller). Coming from my previous 1200 and 2600 that never even got above 80C on full loads, I am very surprised.

Anything I should take a look, or check, other than the usual cooler re-seat, repaste stuff (of which I read is quite decent for a low profile cooler, almost comparable to the Noctua NH-L9i). PBO, Curve Optimizer stuff that I admit haven't really dabbled before? The 5600 should be .. cooler than my previous chips, being more efficient, no?

Granted I haven't try gaming on it, waiting for my 6600xt to arrive.

Thanks for any feedback in advance, guys.

Cooler is insufficient and crap motherboard. Both adds up. Try upgrading the thermal paste to Noctua NT-H2 or Thermal Grizzly Kryonaut Extreme.