SE2: Is fabless MEMS fabulous?

Recent years have witnessed the maturity of MEMS in automotive safety applications and its penetration into cost sensitive consumer markets. With the emergence of consumer applications for sensors and reconfigurable radio, the MEMS industry is at a growth inflection point. Many start-ups are trying to bring low-cost devices to market using a fab-less or a fab-light business model. The pure-play MEMS foundries have grown in number and capability, and cost-effective hermetic encapsulation techniques are now widely available. Despite this, successful commercialization stories are sparse among MEMS start-ups and are mostly generated by the larger and more established companies that have introduced MEMS products after many years of costly internal research and development for high price margin applications. In addition, products have mostly represented single functionality so far while sensor fusion and integration ideas for multi-functionality and reconfigurable chips may become increasingly popular. Here are some questions the MEMS industry is faced with: Are today´s MEMS low cost enough to compete and survive in consumer markets? Given the smaller mask count and larger feature sizes of micromechanical devices, can MEMS die-cost-per-area become less than CMOS die-cost-per-area? What challenges are involved in making this a reality? What integration schemes will have a more promising future? Can planar integration of CMOS and MEMS survive the cost and size scaling? Can we think of a generic process flow that can address 80–90% of the MEMS market?