SAN FRANCISCO Microelectromechanical (MEMS) devices are the next hot market but when they'll flower is anyone's guess, according to panel assembled here for the International Solid State Circuits Conference (ISSCC).
"Thirty-two years ago, I moderated a panel on the growth potential of biosensors, and we are still in its infancy today," panel moderator, Kensall Wise, professor at the department of Electrical Engineering and Computer Science, University of Michigan (Ann Arbor, Mich.), told the audience.
The panel late Monday (Feb. 10) discussed how the IC industry can tap into new markets driven by imagers/displays/MEMS. Essentially panelists asked: Will there be more MEMS on IC chips and will conventional ICs become only a small part of microsystems as MEMS/imagers/displays take away a bigger piece of the system's pie?
"We expect that Aibo (the robot) will be full of MEMS by the time the 2015 RoboCup is held pitting man vs. robot," said Yoshiaki Hagiwara, principal engineer, Sony (Tokyo).
Christofer Hierold, professor at the Swiss Federal Institute of Technology, Zurich, Switzerland, agreed that embedded MEMS will be rage and be marketed as "MEMS Inside" signifying that the product one is buying has value-added functions.
The most vocal promoter of the potential of MEMS was Clark C.T. Nguyen, Program Manager, Microsystems Technology Office, DARPA. He admitted that current MEMS are hard to build and limited in their applications. "But who thought we would have CMOS in the bipolar era?" said Nguyen. "The most successful it becomes the less noticeable it will be."
Nguyen was especially optimistic about MEMS complimenting current electronics: "It should not be a surprise if, in the near future, MEMS and IC technology actually develop a symbiotic relationship in certain applications, where one requires the other for performance advancements." He offered as an example the recent developments where MEMS technologies are used to efficiently cool hot-running CPU chips.
A detractor to MEMS success was Daniel McGrath, an industry consultant in Andover, Mass. "While the opportunity MEMS, imagers and displays is there to provide the key to industry growth, this opportunity will be missed because it requires a different model for the business," he said.
The physical requirements do not scale, he asserted. "Scaling works against performance, and there needs to be a business mechanism to fund relevant device development, design flow and process enhancements," he said.
"You need to have a passion for what you are building," said Roger Howe, professor at the department of Electrical Engineering and Computer Science, University of California, Berkeley, CA. "People will give money if they see that you believe in what you are doing."
Current MEMS devices may be hard to build, reliability is lacking and yields are unacceptable for volume production, but so was CMOS in its first years, said Ray Roop, Director, Strategic Technology at Motorola's Sensor Products Division (Tempe, Ariz.). "The company that finds a way to enhance reliability and yield will control the market," said Roop. "MEMS inertial and pressure sensors have already achieved high-volume production in automotive applications, and will continue to grow in other areas."