First off, I have already read and followed the instructions in the "No POST" sticky thread. In general I found it more aimed at new builds that aren't powering on and have never powered on, whereas my system has been running fine for almost 7 years and suddenly stopped last Sunday.
The system is a Linux machine built by me, running Ubuntu 20.04 as a headless server. I interact with it via SSH and NoMachine only. On Sunday I kicked off a CPU intensive task -- computing a machine learning model -- which was going to take ~8 hours of 100% CPU to complete. I went out for a couple hours and when I came home the machine was not running. It wouldn't boot back up. I connected a monitor to its DVI port and there was no display when powered on.
So I took out all of the components, even the memory, leaving only these parts still connected:
ASRock FM2A85X-ITX motherboard
AMD A10 6700T CPU
Scythe Big Shuriken 2 heatsink
picoPSU-120 power supply
What does happen: The power supply LEDs light up (both the red and green LEDs on the picoPSU), the front panel LEDs light up, the case fan turns on. The heatsink fan runs sporadically, a few rotations, then stops.
What doesn't happen: no display on the monitor, no BIOS screen, no beep on the system speaker.
My first instinct is that I killed the CPU from overheating. The power supply is only 120W, and I have noticed the HSF is often not spinning at all, and when it does it is in fits & starts. I don't know if this is malfunction, a sign of inadequate power, or simply the BIOS dynamically managing fans based on temps. But what I've been reading on other posts is that apparently it is rare for a modern CPU to die from overheating due to the failsafes and protections in place to prevent that. So even if the HSF is unreliable, the system should shut off before the CPU dies.
It seems very strange that I get no beep codes through the system speaker, even though I've removed the memory. Other posts I've read indicate that usually means the power supply is dead. But as I've said, the PSU appears operational since its LEDs come on and the case fan spins.
Any ideas of what I should try next?
The system is a Linux machine built by me, running Ubuntu 20.04 as a headless server. I interact with it via SSH and NoMachine only. On Sunday I kicked off a CPU intensive task -- computing a machine learning model -- which was going to take ~8 hours of 100% CPU to complete. I went out for a couple hours and when I came home the machine was not running. It wouldn't boot back up. I connected a monitor to its DVI port and there was no display when powered on.
So I took out all of the components, even the memory, leaving only these parts still connected:
ASRock FM2A85X-ITX motherboard
AMD A10 6700T CPU
Scythe Big Shuriken 2 heatsink
picoPSU-120 power supply
What does happen: The power supply LEDs light up (both the red and green LEDs on the picoPSU), the front panel LEDs light up, the case fan turns on. The heatsink fan runs sporadically, a few rotations, then stops.
What doesn't happen: no display on the monitor, no BIOS screen, no beep on the system speaker.
My first instinct is that I killed the CPU from overheating. The power supply is only 120W, and I have noticed the HSF is often not spinning at all, and when it does it is in fits & starts. I don't know if this is malfunction, a sign of inadequate power, or simply the BIOS dynamically managing fans based on temps. But what I've been reading on other posts is that apparently it is rare for a modern CPU to die from overheating due to the failsafes and protections in place to prevent that. So even if the HSF is unreliable, the system should shut off before the CPU dies.
It seems very strange that I get no beep codes through the system speaker, even though I've removed the memory. Other posts I've read indicate that usually means the power supply is dead. But as I've said, the PSU appears operational since its LEDs come on and the case fan spins.
Any ideas of what I should try next?
Last edited: