What exactly is causing the bug though? If the same area is written to some x number of times the solid state device in that location permanently fails, is that

What exactly is causing the bug though? If the same area is written to some x number of times the solid state device in that location permanently fails, is that correct? If so in an always on device how can this bug be escaped? They need to randomize the writes and for that the storage size should be a multiple of what is needed for regular operation. Even then the disk will fail eventually. What am I missing here?