As we migrate to Industry 4.0 where we anticipate the next wave of industrial transformations to be led by innovation in autonomous approaches, we have to ask ourselves “What does innovation look like”? To achieve our key [semiconductor] industry metrics of scaling Power, Performance, Area, and Cost (PPAC), it is critical to leverage emerging approaches to data collection, analysis, and modeling, to enable intelligent automation and control. At the same time, as we migrate towards 3D devices, the complexity is increasing to levels that are slowing down time-to-market and increasing costs to unacceptable levels.
From early stages of research and development, through initial yield ramps, and finally to maintaining highly efficient factories during HVM, the common thread is leveraging emerging technology in sensors, big data analytics, computational analysis, virtual models, and intelligent automation to achieve the quickest and most cost efficient path during each of these phases.
As we pursue PPAC scaling, process variability and increasingly stochastic effects, constrain circuit design, preventing us from reaching our targets. An increasing focus on Design-Technology Co-Optimization (DTCO) requires more advanced tools to help us better predict areas we need to reduce variability. At the same time, once a solution path has been identified we need to be able to predict and manage this variability across fleets of tool in a high volume manufacturing environment.
To realize Industry 4.0 we need new tools and most importantly new skillsets in our workforce. Standardization of critical measurement techniques in patterning, ensuring data quality, and data security are just a few of the challenges that we face.