Wierd Failure

I have an odd problem. I have a box that was a Win7 machine, we had some kind of electrical surge at our house that fried a few low voltage components throughout the house, including the Ethernet port on my motherboard. Everything else seemed to work ok, so I got an R45-USB network adapter & kept using the box.
After a short time, I started having lockup issues. I thought it was Windows troubles, so I tried to reinstall, but the computer would suddenly shutdown during the installation. It was like someone unplugged the box.
So I loaded Ubuntu and it works fine, no lockup issues. I really want to use this machine as a Windows box for my wife, she isn't familiar with Linux. I got to thinking that it might be a RAM problem, so I loaded my UBCD and ran memtest86+. It shutdown after a couple of minutes. I removed all of the RAM sticks except for the one in DIMM1, it again shut down after a couple minutes of running memtest86+ and continued to do so for each of the 4 individual 1GB RAM sticks.
Does anyone have any ideas about what else might cause this? I'd suspect the power supply, but everything works perfectly when I'm using it as a Ubuntu system.

 

Sorry, I forgot to give the specs.
It's a MSI P6n SLI mothrboard with 8GB ram, Intel Coe2 CPU 6400@2.13GHz running 64bit Ubuntu 14.04

 

I think you could be right, it could be a failure with the RAM hardware, but not necessarily the RAM sticks, it could be on the board itself and whenever the damaged address it called, it shuts down.

If available, you can test each stick on another DIMM slot or better, another computer.
It would be hard to say what exactly was damaged on the motherboard, and it could be more than your LAN port.

In future, plug your laptop into a surge strip and PC into UPS

 

What about a good old fashion temperature issue. Can you monitor temps and/or fan speeds?
Of course, it may be your power supply. I hope it's not the mobo.

 

You could also to stress your board, under Linux, with a high CPU program like a benchmark.

 

The CPU & GPU temps are fine. I kinda suspect something to do with the mobo (RAM bus maybe?). I may have another board I can use in the box to test the RAM, I just need to dig though all the junk I have on the shelves in my garage.