On 7/1/20 8:01 PM, Thomas Dineen wrote: > I am developing an optimization application that uses the CPU quite > intensely > for long periods of time on a CentOS 6.9 machine. Some test runs can > run for hours or even days. > On a particular test the the OS crashes to a black screen with the > message "North Bridge Disconnected" > printed in the upper left. This list is not for CentOS questions and why are you still on 6.9, which is quite old? > The application is carefully designed to not simply consume too much > memory and run out, and observation > of the System Monitor confirms this. > > What could be the cause? Is this a bug in my code, quite possible > given the fact that I an still testing and debugging? > Can a bug in a user application get into Kernel space and crash the > machine? > Or is it likely a hardware problem on the Motherboard? Unless you're running as root and poking at things you shouldn't, there's no way a user level program should be able to cause that. It sounds most likely to be a hardware issue. Maybe something's overheating?