IBM, Coventor present 22nm virtual fabrication success at SISPAD
September 9, 2013
When will we get 3D NAND Flash Memory???
September 29, 2013The lack of a standard MEMS process has long been bemoaned by those inside and outside the MEMS industry. Standard CMOS processes, after all, have been a key enabler of enormous growth in the ASIC market. If only MEMS could be more like CMOS…
To be sure, the MEMS industry has been making progress. Leading MEMS IDMs like ST, Bosch and Analog Devices have built whole product lines around their respective proprietary processes. Recently, major CMOS foundries like TSMC and GlobalFoundries have announced intentions to offer MEMS processing service. Presumably, they plan to offer standardized processes that will take them on the same path to success that they followed in the CMOS market. For now though, smaller MEMS-focused foundries appear to be leading the way.
In a recent post in the EE Journal, Bryon Moyer describes two efforts, by Teledyne DALSA and CEA Leti, to provide “universal” MEMS processes. These two processes actually address two different challenges. The highlight of Teledyne DALSA’s approach is a wafer scale packaging method that allows fabrication of accelerometers and gryos on the same die. As Moyer explains with his usual technical clarity, the challenge in packaging these two types of devices together is that accelerometers need to be surrounded by gas at atmospheric pressure whereas gyros need to be packaged hermetically in as close to a perfect vacuum as possible. The Leti approach addresses a quite different challenge: how to dramatically shrink the silicon footprint, i.e. the horizontal dimensions, of a MEMS sensor. That benefit is not mentioned in the EE Journal article, but it was my takeaway from hearing a talk by Jean-Philippe Polizzi of CEA Leti at an iNEMI MEMS workshop in March 2012. If successful, the Leti approach may be a game changer because their technical approach of combining nano-scale piezo-resistive sensors with a conventional MEMS proof mass allows a reduction in the horizontal dimensions of the MEMS sensing element from hundreds of microns on a side to tens of microns.
While standardized MEMS processes like these two will no doubt gain traction, the extent to which either one is universal can be called into question. The Teledyne DALSA and CEA Leti processes appear well suited for inertial sensors (accelerometers and gyros), but I would hazard an educated guess that either process will require substantial modifications for microphone designs, or other types of MEMS. As Bryon Moyer points out, the lack of standardization is not a bad thing, or even undesirable. He notes, “…in the absence of an enforced standard, innovative ideas have a better chance of seeing daylight.”
The key points, with or without standardization, are 1) there will continue to be diversity and innovation in MEMS processes, and 2) MEMS design and MEMS process development will continue to go hand-in-hand to a much greater extent than has so far been the case for CMOS. Traditionally, MEMS companies have relied on long series of build-and-test cycles to address these challenges, with attendant lengthy delays in bringing products to market. With increasing competition and time-to-market pressure, MEMS companies must become more efficient. That’s where Coventor comes in. We believe that our software solutions can unlock a more efficient way innovate, one that is in sync with the cost and time realities MEMS companies are facing. We’ve committed ourselves to working closely with our customers to make this happen.
![Fig. 2. Silicon spin qubit centered in the dotted circle, control and readout signals (M, P, R, T, and Q) are shown in the inset. Simplified schematics of the quantum point contact and corollary circuits are shown. The voltage source is implemented as a digital-to-analog converter at room temperature. (Taken from [6])](https://www.coventor.com/wp-content/uploads/2022/05/Fig.-2.-Silicon-spin-qubit-centered-in-the-dotted-circle-control-and-readout.png)
![Figure 2: Schematic view of standard CMOS including an electro-mechanical switch monolithically integrated into BEOL [2]. Courtesy: University of California, Berkeley](https://www.coventor.com/wp-content/uploads/2022/04/Figure-2-April-2022-MEMS-Blog-for-Coventor-website.png)
