Kernel panic when booting past BIOS

[jlicht]

Hi guys,

I'm having some pretty serious issues with my very new rig, and I was hoping someone here might be able to offer some insight into what the problem could be.

I was using my computer as normal, when suddenly the computer switched off.
My CPU temperatures are generally between 30 and 70 degrees and my GPU peaks at 83 degrees, so I don't think this was temperature related.

Now my computer can boot into BIOS, but kernel panics if I boot into anything else.
See for example trying to boot into Windows 10 on internal SSD in the video found in this folder called bsod.m4v:
https://www.asuswebstorage.com/navigate/s/01F249AF13704CFFB588B7F233DA09464

I have the setup listed here: https://pcpartpicker.com/list/CBLDd6 with an ASUS Z270i and a Core i5 7600K.
There are also pictures of my setup in the folder above.

In order to pinpoint the problem, I have tried the following things, all of which resulted in the same problem:
- Booting Windows 10 from the SSD (BSOD, freeze or instant reboot)
- Booting Windows 10 installation media (BSOD, freeze or instant reboot)
- Booting from Ubuntu Live USB (kernel panic, see screenshot ubuntu_live_usb.png)
- Booting from Debian live USB (kernel panic)
- Accessing a shell from Ubuntu USB (kernel panic, see screenshot ubuntu_shell.png)
- Plugging out the SSD
- Using different RAM
- Disabled memory remapping, audio controller, wifi controller, network controller in BIOS
- Taking out the GPU and using onboard graphics

The above leads to me to induce that there is a serious hardware problem in either the motherboard or the CPU.

Both my motherboard and CPU were bought in May, so they are very new.
If one of them has a problem, I can send it to RMA. But first I need to determine which part has the problem.

Does anyone have any clues to where the issue is, or what I could try to further narrow down the issue?

Cheers,
Johannes

 

[molletts]

Hi Johannes,

That's a tricky one - I've had a few like that in the past and they're not easy to debug if you haven't got access to an alternate CPU to try in the motherboard and another board to test the CPU in. (I even bought a cheap Athlon II on one occasion just to try and rule out a faulty CPU as the cause of problems I was having with the PC I was using at the time - it's in my home server now.)

I'm guessing you've tried booting Ubuntu with different RAM, no GPU and no SSD all at the same time (i.e. just the minimum hardware needed to boot the USB key) and that you're not overclocking (or have reverted to stock speeds)?

Have you tried a different PSU (if you have access to one)? Perhaps its output voltage is unstable, which could cause the system to behave strangely, especially as the amount of power drawn can change suddenly as the OS initialises devices, etc. during boot.

Other than that, your best bet might be to ask ASUS Tech Support for help - they may have some kind of diagnostic tool that you can run to test the motherboard and/or CPU, or they may simply say that the best thing to do is to swap the board and if that doesn't solve it, it's probably the CPU.

Sorry I couldn't give you anything more specific

Stephen

 

[t53186]

Wouldn't hurt to flash you bios, put in minimum components needed to boot (CPU, min ram, no drives) , enter bios and verify config, then boot to a usb. Start installing drives, additional ram, and if you still get the same result rma both.

 

[jlicht]

I'm guessing you've tried booting Ubuntu with different RAM, no GPU and no SSD all at the same time (i.e. just the minimum hardware needed to boot the USB key) and that you're not overclocking (or have reverted to stock speeds)?

Have you tried a different PSU (if you have access to one)? Perhaps its output voltage is unstable, which could cause the system to behave strangely, especially as the amount of power drawn can change suddenly as the OS initialises devices, etc. during boot.

Hi Stephen, thanks for your time 🙂 yes, I tried booting both operating system with different RAM, without the GPU and with no SSD. Both CPU and RAM are set to stock speeds.
I can try with a different PSU just in case. I'll try to ask around work for a spare one.

Wouldn't hurt to flash you bios, put in minimum components needed to boot (CPU, min ram, no drives) , enter bios and verify config, then boot to a usb. Start installing drives, additional ram, and if you still get the same result rma both.

BIOS is already the newest version. I tried booting with a minimal configuration and the problem is the same :-/
I'm just a bit afraid that I will be charged for whatever component is not faulty. Luckily I at least bought both CPU and motherboard from the same retailer.

Would you go through Intel/ASUS directly or through the retailer? I already opened technical support tickets with both Intel and ASUS, but I'm not sure about the RMA itself.

 

[t53186]

The retailer for the warranty. Check their return policy, I had a similar problem and sent both cpu and MB back to Newegg without a problem.

 

[jlicht]

ASUS support's answer was:
"I think it looks like a broken Mainboard. Please contact your vendor for an exchange."

Intel support's answer was that DHL will come pick up my CPU on Tuesday and send me a new one.

I suppose that's pretty expedient service, although it seems a little silly to replace both components when probably one of them is faulty :-/

 

[t53186]

ASUS support's answer was:
"I think it looks like a broken Mainboard. Please contact your vendor for an exchange."

Intel support's answer was that DHL will come pick up my CPU on Tuesday and send me a new one.

I suppose that's pretty expedient service, although it seems a little silly to replace both components when probably one of them is faulty :-/

Unless the MB damaged the CPU, then if all you got was a new cpu it would become toast as well.

 

[molletts]

Well, hopefully replacing both CPU and motherboard should fix the issue. If it doesn't, at least it rules them both out, although it would leave you with a much more puzzling problem to solve!

I think I've only ever had one faulty CPU in 25+ years, which I spent ages troubleshooting, swapping in known-good parts one at a time from another identical server whenever I could get a short period of downtime. (Apart from a very old system where it wasn't worth the time, effort or money to troubleshoot it and speculatively replace parts to determine for sure whether it was a CPU or MB issue.)

 

[jlicht]

Well, hopefully replacing both CPU and motherboard should fix the issue. If it doesn't, at least it rules them both out, although it would leave you with a much more puzzling problem to solve!

I think I've only ever had one faulty CPU in 25+ years, which I spent ages troubleshooting, swapping in known-good parts one at a time from another identical server whenever I could get a short period of downtime. (Apart from a very old system where it wasn't worth the time, effort or money to troubleshoot it and speculatively replace parts to determine for sure whether it was a CPU or MB issue.)

Yeah, I also find it unlikely that the CPU was the cause of the problem.

While waiting for my new parts, there's one more thing to consider, namely the PSU.
I've been googling around a bit, and it doesn't seem likely that undervoltage would have damaged anything, but it is a fairly small PSU (450W).

Do you think there's any point in my trying to sell it and upgrade to a 600W, or would that just be silly? ;-)

Once I get the new parts I can put a power meter there and see what the actual consumption is, but I of course don't want to damage anything a second time...

 

[molletts]

While waiting for my new parts, there's one more thing to consider, namely the PSU.
I've been googling around a bit, and it doesn't seem likely that undervoltage would have damaged anything, but it is a fairly small PSU (450W).

Do you think there's any point in my trying to sell it and upgrade to a 600W, or would that just be silly? ;-)

Once I get the new parts I can put a power meter there and see what the actual consumption is, but I of course don't want to damage anything a second time...

I think 450W should be big enough, albeit without much headroom. I must admit that I would have chosen a slightly beefier one just to be on the safe side and allow for future upgrades or possible overclocking. When you put the system on a power meter, though, you'll probably find that it doesn't use much more than about 250W even under heavy load, except when running benchmarks specifically designed to max out various components. (That, of course, is the total overall power draw; it's a bit trickier to measure the draw from individual rails of the PSU unless you have one of those fancy "smart PSUs" with built-in power metering.)

The SF450 appears to be able to deliver up to the full 450W on +12V so it should be OK. The most likely thing to happen if a (good) PSU gets overloaded and the voltage starts to drop is that it will either shut down or de-assert the "power good" signal, which will normally cause the system to restart.

Did you manage to test the system with a different PSU in the end?

 

[jlicht]

I think 450W should be big enough, albeit without much headroom. I must admit that I would have chosen a slightly beefier one just to be on the safe side and allow for future upgrades or possible overclocking. When you put the system on a power meter, though, you'll probably find that it doesn't use much more than about 250W even under heavy load, except when running benchmarks specifically designed to max out various components. (That, of course, is the total overall power draw; it's a bit trickier to measure the draw from individual rails of the PSU unless you have one of those fancy "smart PSUs" with built-in power metering.)

The SF450 appears to be able to deliver up to the full 450W on +12V so it should be OK. The most likely thing to happen if a (good) PSU gets overloaded and the voltage starts to drop is that it will either shut down or de-assert the "power good" signal, which will normally cause the system to restart.

Did you manage to test the system with a different PSU in the end?

Intel was so fast in sending a courier that I had to quickly disassemble everything, so I didn't get a chance to. However, if the problem was related to the PSU it would probably have been the cause of whatever damaged the motherboard, not directly cause the kernel panics (that would be very odd).

I ran the 7600K at 4.70 GHz for about a month until the incident, so the PSU seemed to be able to keep up with the overclocking at least for that period.

You're right that the overall wattage is probably not the thing to look at, so might not be much I can do there.
Only thing is that if I do want to replace my PSU, I'm better off selling it now, as I could still sell it on as practically brand new with five years warranty 🙂