Assessment of microelectronics interconnect reliability - current practice and trends

When it comes to the long-term reliability of the most common microelectronics interconnects, too little effort is spent on investigating important aspects while a significant amount of day-to-day reliability testing may be more or less wasted. This would seem to be at least partially so because reliability managers often fail to ask themselves what it is that they really want to know. Notably, while rarely explicitly recognized even ongoing `engineering tests´ will usually be meaningless unless they reveal something about relative performance in service. This requires a model or at least a quantitative mechanistic understanding.