Can Linux kill a motherboard?

[elijahk33]

I was using Ubuntu 14.04 on my Asrock B85M-ITX when the motherboard just died. It died! No fans, no leds, nothing. I tried all the common troubleshooting passages: trying another psu, 1 ram per time, everything outside the case. Nothing.
Do you think linux killed my motherboard (or cpu?????)?

[Kano]

Most likely you can revive it with a freshly programmed firmware chip. Some boards can flash even without CPU from USB, but maybe a bit too cheap for that feature. In the case you own a board with flashrom support you can try hotswap.

[elijahk33]

I thought about that too. Unfortunately Asrock has removed the USB recovery feature from its AMI bios chips. So no recovery flash.
I've ordered a new ITX board from Asus. I hope nothing else has broken.

Is this something that happens very often?

[Kano]

Just send it back if you still habe warranty. It is basically not Linux specific but you should update the firmware if possible. UEFI is relatively new and different to BIOS. There is no real CMOS clear possible, now this is a software function, all settings are stored in the firmware chip itself. So a reset is not possible if it gets too currupted.

[cryptoahash]

Do you think linux killed my motherboard (or cpu?????)?

lol sure. Linux is a very dangerous beast that will eat all your hardware.

[souenzzo]

Motherboard ??
It seems that this error has not been reported
Sometimes it eats your RAM ...

http://www.linuxatemyram.com/

[Luke]

This sounds like a sudden hardware failure

I was using Ubuntu 14.04 on my Asrock B85M-ITX when the motherboard just died. It died! No fans, no leds, nothing. I tried all the common troubleshooting passages: trying another psu, 1 ram per time, everything outside the case. Nothing.
Do you think linux killed my motherboard (or cpu?????)?

Most likely this came from a sudden failure of a component, both the board and the CPU are suspects, but first thing to check is your power supply and your power supply connections(which may have been loose). Disconnect all PSU leads, then with the PSU plugged in short the GREEN lead momentarily to a BLACK ground lead. The PSU should start and its fan should run unless it's the kind with zero fan speed at light load. In that case you might need to put a voltmeter to a 12V white or 5V red lead to verify operation. If the PSU does not start, check the AC power cord! There are lots of power cords that make poor contact to the AC pins on the outside of the PSU, this is like a car being out of gas or having a plugged fuel filter.

Next plug the 24 pin connector and the 4 or 8 pin "Pentium 4" connector back into the board, making sure to seat them fully but not to push so hard as to damage the board. With the PSU plugged in, try the power button. No response at all usually indicates dead components on the board. I would expect a firmware issue to act more like a Pentium 4 board I had that came with a dead CPU on a good board: No POST, fans to full power or at least running. The ugly part is that a sudden failure of the power circuits on the board (I've had that), can also destroy the CPU as well. If the board won't start on a known good PSU, than consider it dead and take it back assuming it has never handled encrypted data held against law enforcement or against the manufacturer of the board. In any case you will need another one.

I've not once had clear CMOS start up a BIOS board that was refusing to POST for any reason other than pushing BIOS overclocking too far!

In the only known cases where a board was truly incompatable in UEFI mode with Linux in a "bricking" way, the board was "bricked" as soon as you tried to boot if and only if you tried to boot in UEFI (not "legacy" mode). This was due to a severe UEFI bug, forget who made those laptops but Windoze didn't invoke the bug. I think there were Linux kernel changes to also avoid invoking the bug, but the bug was in firmware and neither Windows nor Linux should have had to resort to workarounds.

Even then, if those laptops contained any kind of slot giving DMA access it would have been possible to use any kind of "bus pirate" device that runs its own CPU and becomes bus master to forcibly reflash the firmware and to rewrite whatever value in flash storage had become corrupted, assuming that value was known and a firmware image was available. Of course, if your motherboard has a hardware issue, reflashing the firmware won't do anything, assuming the damaged hardware would even let enough of the board run to power the flash chips. I would not waste time and money getting and learning to use the Bus Pirate when hardware failure is the main suspect. If Linux was responsible the issue would have appeared at first boot, or hypothetically at first run of some offending program that invoked low level commands not normally used, like an overclocking from the OS program might have to do.

[elijahk33]

I got the new motherboard (Asus H81I-Plus). Now the pc is working fine.
So the problem was the Asrock B85M-ITX motherboard which died suddenly. I still don't know why. I've checked the motherboard visually: no burn marks, no socket pins bent.

[System25]

I think it is unlikely Linux could cause complete MB death. Complete death usually happens due to hardware or firmware issues. While in some cases it is possible to kill hardware via software methods, it is uncommon for desktop MBs, except maybe erased BIOS (and it's not like if Linux does something like this during normal course of actions).

And if someone really curious about death cause of MB, you'll have to check power supply 1st - voltages and if these are in valid range, as well as if PowerGood signal is present and if MB actually activates power supply upon ATX button press. If everything is okay (i.e. main supply is activated, but no proper startup), you may want "POST codes card" (usually small PCI or PCE-E PCB with 7-segment leds). If hardware is at least partially working, you'll see how early startup progresses and then last code where it stuck, indicating what BIOS did at this time. This code could give idea what goes wrong if bios always getting stuck at same point. Still depends on what exactly died though. And as you can guess, existence of POST codes boards suggests you're not first human on that planet who got MB dead. But well, if you're not good in electronics, it is better idea to contact experienced technician who knows how it works and owns right equipment to carefully check troublesome points.

[A Laggy Grunt]

I doubt it.

I did have a P4 laptop (Dell Inspiron 9100). The people at Dell put a desktop Prescott in that thing. One day, the fans didn't spin up when I applied load, and... poof. No fire, just something died. I suppose it's possible if something mismanages your fans and one of those thermal safety type mechanisms aren't triggered.