I'm trying to install Ubuntu on my new to me HP prodesk 600 g4. It's second hand but from a refurb company so it should be in an ok state hardware wise.
The first time I noticed issues with the install was when the keyboard taps wouldn't respond during the install process from a live USB boot. I thought it was my keyboard's fault but I use this keyboard all the time and never had an issue. After forcing a hard reboot each time I did this the system would lock up at progressively sooner points in the process, often not even getting to the install wizard.
I managed to get it installed, but it locked up almost straight away on first boot.
I reflashed the install media just in case it was corrupt, tried again trying to tap through the wizard as fast as I could to avoid lockups, and it still did it.
I disabled secure boot and tried again, no change.
I reset the BIOS settings, re-disabled secure boot and tried again, no change.
I hypothesised that it was the CPU maybe locking up from inadequate cooling, but I managed to install sensors package and it's idling at 37-39 Celsius.
I've even tried Ubuntu 20.04 just in case it was a new problem, and the same thing happened after first boot.
The box came with windows installed, but I never gave that a go because I never had the intention of using it so I don't know if the issue was there too.
I have done a full memtest86 and it showed absolutely no errors at all.
I thought I had a breakthrough when I tried taking out one stick of ram and tying again, ignoring the memtest and it miraculously seemed to run fine for a few hours! Wow, it might be ram I thought. I swapped the stick and slot and it still worked fine. So I put both sticks back in and tried again and it ran for several hours overnight. Strange, maybe the ram just needed to be reseated?
This morning I went to check on it and nope, it's locked up again.
When it locks up there are no errors, no warnings, no oddness in the screen or SSH session. It literally just pauses and the cursor on screen stops blinking.
I checked the last logs with journalctl -b -1 -e and it showed one error: i915 0000:00:02.0: [drm] *error* cpu pipe a fifo underrun. I have no idea if this is related but searches online don't seem to show the same behaviour. People with this seem to experience graphical glitches but the computer keeps running.
When booting live OSes from ventoy it seems to run fine. I haven't tried many options but hirens bootcd seems to work with no issues and memtest86 ran all the way through without locking up so I think it's Ubuntu 22.04 or kernel specific but I don't know enough to be sure.
I've also seen some theories about NVIDIA drivers but to my knowledge this doesn't have an NVIDIA GPU in it so I don't think it should be that?
Please any help would be greatly appreciated.
After the most recent freeze I looked at journalctl again and that same error wasn't even there that time. I'm not sure it's related, I've only seen that once. I can't seen any other relevant warnings in journalctl. There is also absolutely nothing in /var/crash. It's like it just freezes in time.
I have verified the ISO is valid as per one suggestion also. It is valid.
The next thing I might try is unplugging the HDMI cable and letting it run headless. I only have HDMI on the HP FlexIO port and haven't used the other displayport ports.
Running headless didn't solve it
