From the MEMS Investor Journal / MEPTEC MicroPower Workshop

Paul Werbaneth, Vice President Marketing & Applications, Tegal Corporation

From the MEMS Investor Journal / MEPTEC MicroPower Workshop, Thursday 22 October 2009, San Jose, CA:

Hot or cool. Fire or ice. Temperature differentials across a Peltier device.

The 1^st Annual MicroPower Workshop, put on by the MEMS Investor Journal (great job Mike Pinelis!) and MEPTEC (great job Bette Cooper!) couldn’t have started stronger, or ended stronger, or been too much better in between.  To start:  Professor Al Pisano, UC Berkeley, speaking on “Multi-Fuel Micro Engines.”

MEMS at the Gartner Semiconductor Briefing

Paul Werbaneth, Vice President Marketing & Applications, Tegal Corporation

I was fortunate to catch Jim Tully, VP, Distinguished Analyst, Semiconductor Research, Gartner, speaking on “Emerging Technologies in Semiconductors and Electronics” at the Gartner Semiconductor Briefing held at the Doubletree Hotel in San Jose on Thursday morning, October 15, 2009. (Wonderful breakfast, courtesy of Gartner, by the way, me sitting at random next to the very droll Chanan Greenberg from the company Model N; nothing like wry and dry British humor with your morning kippers.) The key issue for Jim on the (now cleared from breakfast) table: which technologies will drive growth in semiconductors over the next 5 – 10 years?

First, it’s 3D wafer stacking, which climbed and descended the formidable peak of inflated expectations, crossed the difficult trough of disillusionment, and now, with the risen sun steadily evaporating what remains of early dew, can be seen well up the slope of enlightenment, on its way to the plateau of productivity. Is 3D wafer stacking MEMS technology? It’s not a point Jim explicitly makes, but I want to make it for him – without the deep supply chain background of wafer bonding from MEMS, and silicon DRIE from MEMS, 3D wafer stacking would still be looking up at the summit ridge on Mt. Inflapectation, and probably be wanting to poke around one side or the other for easier ways to cross. Thanks to all those fixed routes left by MEMS parties, the best route now is direttissema.

And what else? Jim says there are MEMS mirror-based picoprojectors on his list. The underlying technology must be MEMS fabrication I think, since picoprojectors are a product, at least to me. “Biochips and electronics are merging” Jim says, showing as examples a bird flu Lab-on-chip detector from STMicroelectronics, as well as a scary looking neuron-on-silicon light sensing device from IMEC. MEMS-enabled SmartPills, MEMS-enabled cell phone handset HealthOMeters that sample sweat from a user’s hand and provide up-to-the-minute data on body chemistry and condition (and maybe are encoded to unlock stored-value payment accounts by responding only to the registered user’s DNA – trying cracking that safe Alexander Mundy).

To be sure, Jim thinks “further integration of MEMS, image sensors and wireless technology will open up new application areas in games, microphones and many other areas.”

And how to power these now and future wonders? “Micro fuel cells offer ten times the energy storage of lithium ion batteries. Energy scavenging schemes involve various technologies that convert energy in the environment into electricity. Micro turbines are a micro-electro-mechanical system (MEMS) technology that would implement a tiny internal combustion engine within a package about the size of a semiconductor chip.”

Jim, you’re a semiconductor guy, but you are most welcome at the MEMS Executive Congress next month. We couldn’t have been any better champions of MEMS commercialization than you were this Thursday at the Gartner Briefing. We’ll be discussing just about everything you identified and like about the MEMS-enabled future when we convene next month in Sonoma. See you there?

MEMS Microphones Enjoying Dynamic Times, Resurgence in Attention

By Jérémie Bouchaud, Principal Analyst, and Richard Dixon, Senior Analyst, MEMS, iSuppli

Introduction

MEMS microphones are under the spotlight this year with two major acquisitions in the last few months. In May, EPCOS bought Pulse Engineering, while just a few weeks back in August, Bosch acquired Akustica. Despite the recent attention, MEMS microphones are not really new. Knowles, an acoustic component manufacturer also producing conventional Electret Condenser Microphones (ECM), pioneered silicon MEMS microphones 20 years ago initially (with R&D started in 1988) for high-end applications at NASA. The company started to ship to cell-phone companies in 2003. Volumes have grown very rapidly since then, and Knowles announced in August its 1 billionth MEMS microphone (see News section).

Not surprisingly, Knowles is not alone, and iSuppli has identified nine companies currently shipping MEMS microphones to customers and another three sampling or in the R&D phase. Will the market be big enough to feed these companies? In answer to this, iSuppli has just published a report on MEMS microphones, the findings of which we summarize in this issue’s MEMS market brief.

Figure 1: Silicon MEMS Microphone Market

Market Overview

The market for MEMS microphones was in the range of $135 million in 2008 and will more than double in revenue by 2013 to exceed $300 million. By this time, more than a billion units will be shipping annually.

The main applications today and in the future include cell phones, far in front of laptops, and headsets in third place. iSuppli’s Teardown Service also found MEMS microphones in other consumer electronics equipment, such as video cameras (see Teardown of the Month section), and MEMS microphones will also find their way in the near future into ultra-thin DSCs and possibly MP3 players. Noteworthy, suppliers like MEMSTech and Omron also serve the industrial and security markets. In addition, iSuppli has investigated automotive and medical applications, but contrary to other what analysts say, we do not believe that MEMS microphones will ship in this sector in series by 2013.

First Bump in the Road for MEMS Microphones

MEMS microphones have not had a completely smooth ride, hitting a bump in the road in 2009 , as a result of Motorola’s freefall in the mobile handset market. Motorola was the first company to widely adopt MEMS microphones for its mobile handsets, especially the best-selling ultra-thin RAZR, and has adopted MEMS microphones for almost all its phones.

With Motorola dropping from its place as the No. 2 handset manufacturer in 2006 to No. 4 in 2008 and continuing to lose market share in 2009 (source: iSuppli “Wireless Handset Market Shares through Q2 2009”), this has significant implications for the development of the MEMS microphone market. Through numerous iSuppli teardowns, we have been able to estimate the penetration and consumption of MEMS microphones at the top cell-phone makers. The significant growth of MEMS microphones at other handset makers cannot compensate for Motorola’s decline, as Figure 2 clearly indicates.

If one were to superimpose on this the lower overall shipments of cell phones this year by 12% (source: iSuppli Wireless Systems Q2 2009 Market Tracker), as well take into account the significant price erosion this results in a close to modest growth of the overall MEMS microphone market in units and 2% decline in revenue this year.

Figure 2: Motorola’s Freefall Impact on Shipment of MEMS Microphones in Mobile Handsets

MEMS vs. ECM

MEMS microphones still sell at a higher price than the traditional ECM microphones, from less than $0.35 for analog microphones in high volumes at cell-phone companies to more than $0.80 for digital MEMS microphones selling in high-end acoustic equipment. What justifies this price premium?

If you ask MEMS suppliers, some will describe the ECM as an outdated technology and will start off on a long list of decisive advantages imparted by MEMS. Another group of ECM detractors, consisting of MEMS microphones developers, are in the same camp—even though they conveniently neglect to mention that ECM has not been standing still in the last few years. In this regard, iSuppli asked electronic equipment makers (i.e., leading cell-phone makers), both in the acoustic purchasing and research department, to furnish us with their views.

In essence, being an SMD (surface-mount device) is no longer a true Unique Selling Point (USP) for MEMS. There is no decisive advantage that is true for all possible customers, and each OEM will have its own motivation for choosing to implement MEMS instead of ECM technology. Some choose the form factor delivered by MEMS microphones for integration in ultra-thin phones or cameras, and some do so because they want a digital output; some will favor MEMS microphones for their fast ramp-up (or ramp-down), and yet others will make the choice based on performance.

Motorola RAZR gave momentum to MEMS microphones

Motorola RAZR exploded view

Teardown showing Knowles MEMS microphone (Source: iSuppli Teardown Analysis)

Is the Future Digital?

In fact, MEMS offers a better technology alternative for microphones with digital output than does ECM. This is the reason digital MEMS microphones from Akustica and Knowles are so popular among laptop manufacturers. In laptops the audio signal transits behind the LCD screen where there is strong EMI. A digital MEMS microphone is immune to this interference.

If MEMS microphones work beautifully with laptops, can they do the same with phones? Some trends favor the use of digital microphones in future mobile handsets:

• Increasing EMI issues in some handsets as the antenna is buried deeper in the device or as smart phones become more complex.
• New functionalities, such as noise cancellation and beam forming, can benefit from digital signals.
• A standard interface is emerging—SLIMBus—that should ease implementation of digital acoustic signals.

OEMs remain very comfortable with analog signals, though. Digital MEMS microphones were found by the iSuppli Teardown Service only in Nokia’s handsets, while all the other handsets examined by Teardown used analog microphones. iSuppli believes that the share of digital microphones in handsets will be modest until 2012.

Ultra-small-size analog MEMS microphone (3.76 x 2.95 x 1.1 mm, KRM0300), microphone transducer, and ASIC from Infineon (Source: Hosiden)

The Second Source Issue

At present, MEMS microphones are shipped from nine companies. While this sounds like a flourishing trade, the reality is markedly different. The market leader—Knowles—will account for 85% of the market in 2009, leaving the other eight far behind. Last year, no other company exceeded 7 million units. The fact that one company dominates the market so overwhelmingly is an issue for cell phones, as some OEMs have no wish to rely on one source and are therefore reluctant to switch to MEMS microphone technology.

This year, Infineon and MEMSTech are finally emerging from the mass of small players as second and third sources in significant volume, and both are currently expanding production capacity. Noteworthy, these two companies are partnering with leading ECM companies for the commercialization of MEMS microphones.

What will the supplier landscape look like in 2013? One must differentiate from the high-volume supply in cell phones—with low margins—competing against the ECM on price on the one hand, and against the suppliers on the other hand for higher-end equipment that compete on performance (e.g., the emerging companies ADI and Wolfson). Generally speaking, these observations can be made:

• There is room and enough segments on the higher end of the market for three to four players.
• On the lower end, it is not clear whether a second supplier will be successful with fully packaged MEMS microphones to compete along with Knowles against the ECM makers. iSuppli believes, however, that an interesting opportunity exists for MEMS companies to supply microphone dies and ASICs to ECM companies that can package the microphones and sell them by using existing networks. This is the model deployed by Infineon today with Japanese ECM maker Hosiden. Wolfson has also recently started doing the same with other ECMs companies. This way, MEMS companies benefit from the existing brand and sales channels of the established ECM companies and hope to keep higher margins, selling silicon and shrinking die sizes.

iSuppli has just published a thorough analysis of the microphone market and supplier landscape in a 100-page report. The documents include detailed units, ASP revenue by application, and type of microphone (analog versus digital), as well as the profiles of 14 suppliers with their status of commercialization and suppliers’ agreement. Also included is a database of over 75 end products (cell phones, laptops, and cameras) that feature MEMS microphones.

At IEEE 3D IC 2009, Day 2 and 3, San Francisco, CA …..

Paul Werbaneth,Vice President – Marketing and Applications, Tegal Corporation

Thoughts on Day 2 and conclusions following Day 3:

There’s been a really strong student / university contribution to the conference this week by many of the schools to which I feel close.  Penn State, Pitt, Tohoku University. (I had the unexpected pleasure of choosing a lunch table Wednesday where I sat next to Professor Koyanagi from Tohoku University, first time for me to meet this distinguished scholar, who is an IEEE Fellow and an IEEE Medal Laureate.)  Rensselaer (Prof. Ken Rose is my old friend from ASMC committee work, and my niece is now a freshman at RPI), MIT, Nanyang Technological University (Singapore; Prof. Chuan Seng Tan is assembling a book on 3D IC Technology, to which I contributed a chapter on TSV etching), NC State.  Among others.

Kelli Ireland, from Pitt (“A routerless system level interconnection network for 3D integrated systems”) and Adam Beece from RPI (“Impact of parameter accuracy on 3D design”), grad students both, absolutely nailed their presentations – great job Adam, great job Kelli.

When you take someone from Singapore (best food in the world) to dinner in San Francisco where do you go?  I figured since we had my colleague Yannick Pilloux with us it should be Californian French, so off to Grand Café (http://www.grandcafe-sf.com/) we go.  It works for Yannick (“authentic boef bourguignon”), it works for Chuan Seng (Petaluma duck breast), and it works for me (porc roti with mushrooms, beans, and corn).

If Day 2 of IEEE 3D IC was devoted somewhat to Comp Sci topics and Networks on Chips, then Day 3, Wednesday, swings back to devices, applications, and markets.  Eric Beyne (IMEC) pinch-hits and scores with the talk “European R&D activities for 3D integrated heterogeneous systems.”  To me, it really looks like the commercial products we’ll be seeing, enabled by 3D IC technology, are going to contain MEMS stacked onto logic and power, with a radio tucked in somewhere along the way.  At least that’s what they’re saying in Europe, and that’s what (“10um fine pitch Cu/Sn micro-bumps for 3-D super-chip stack”) they’re saying in Japan.

It makes me wonder where Apple is on all this.  IEEE 3D IC is all about great technology, and Apple is all about great (insanely great) products – products based on the best cutting-edge technology and software, of course.  When I start thinking about sleekly designed consumer products packing the power of a complete consumer entertainment system, and boasting wireless connectivity, GPS navigation, and ripping games, all in a single unit,  I look to my iPhone.  Already highly integrated, but then I think, what if those many digital, analog, power management, RF, and MEMS components were stacked into a single packaged “chip,” not spread out over business card-size circuit boards, as they are today.  What more would I get?   (What more do I need?)

How about some health monitoring functions (room for that once we integrate the super dreamy chip into the handset), how about some real-time language translation functions, how about a microprojector, how about a portable augmented reality function that reports to the display glasses I’m wearing (Oak-Bans?) using GPS data so it knows where I am and how my reality should be augmented?  They say there might be room for a printer in there too once we free up the space.

A 21^st century electronic Baedeker, that’s what I want.  With a beautiful red leather cover, just like in the old days.  (I’m still going to be holding this DreamyChip Baedeker in my hand, by the way – direct brain implants are for IEEE 3D IC 2019).  I’m not alone in that I see this coming, propelled by the work reported on this week during IEEE 3D IC.

From San Francisco, CA, thanks for reading.

At IEEE-3D IC 2009, San Francisco, CA

Paul Werbaneth,Vice President – Marketing and Applications, Tegal Corporation

Just back from a week hiking in Yosemite (thank you John Muir!) and in the Bridgeport – Twin Lakes – Matterhorn Peak area (thank you Gary Snyder/ Jack Kerouac and all past-present-future dharma bums).  Bodie, CA, an abandoned gold mining town at the end of the world (at the end of a three-mile washboarded dirt road, really, off a twelve-mile country road, off CA 395) gives literal proof to the phrase “a flash in the pan,” having been a town of ten thousand upstanding, barely standing, or once-standing citizens only 125 years ago.  What’s left now, with the gold run out, is a collection of highly weathered wooden buildings (most barely standing), a book’s worth of stories great, ghostly, and sad, and a permanent population of zero.  Monument to an American Ozymandias, installed and forgotten over the short span from the time of my grandparents’ births to today:

“Look on my works, ye Mighty, and despair!”

Nothing beside remains.  Round the decay

Of that colossal wreck, boundless and bare,

The lone and level sands stretch far away.

(Thank you Percy Bysshe Shelley.)

But not so here at the Hyatt Regency Embarcadero in San Francisco, where the talk is all about 3D System Integration (www.3dic-conf.org/) .  Now you might have been led to think the papers and discussions today would have all been about 300mm CMOS Logic and NAND Flash stacked integration using through silicon vias, but almost all the papers today (OK, about 75% of the papers) talked about integrating MEMS components with logic, and other functions, using 3D techniques (including TSVs, in situ, or TSVs in interposers), in order to create something new.  Something disruptive.  Something super smartphone. (Just what is the right label for what follows “smartphone”?)

Your supersmartphone will likely have a Dream Chip (Morihiro Kada, Association of Super-Advanced Electronics Technologies, “Development of functionally innovative 3D-integrated circuit technology”) on-board, doing what it is Dream Chips do, perhaps brought to wide-spread commercialization by Qualcomm (Matt Nowak, “High density silicon stacking – how the fabless supply chain impacts technology decisions”).  Something more from Japan, from Tohoku University, hot-bed of MEMS development:  “Heterogeneous integration technology for MEMS-LSI multi-chip module” (K.W. Lee).  (I’d keep my eye turned toward Sendai based on the data here.)

A very interesting paper from the University of Pittsburgh, “3D integrated circuits for lab-on-chip applications (Samuel Dickerson) pulled together MEMS, 3D integration, and the medical research expertise for which Pitt is so well known.  Just waiting for MIT Lincoln Labs to send back the fabricated chip so Pitt can test it.

And, shades of Project Sun SPOT (MEC 2008), Philips (Ric van Doremalen) told us today about a “Miniature wireless activity monitor using 3D system integration.”  It’s got a radio, it’s got a battery, it’s got an antenna, logic, integrated passives, and it’s got a MEMS accelerometer, integrated into about one cubic centimeter of total space.  All for a healthier, better, wirelessly connected you.