MEMS likely to see strong recovery in 2010!

Recently, Jérémie Bouchaud – Director and Principal Analyst, MEMS, iSuppli, ran a program on High-margin MEMS: Medical, Industrial and Aerospace.

According to Bouchaud, the MEMS market is not immune to the downturn and is down 8 percent in 2009. The bottom was hit in Q1-09. However, the MEMS segment is likely to see strong recovery and regain double digit growth in 2010.

The high end markets enjoy similar growth as consumer and mobile MEMS markets. It has grown at 16 percent CAGR from 2008-2013. There are opportunities for small companies and foundries. Volume manufacturers have also come to high end markets.

Overview for high-margin MEMS
Providing a market and suppliers overview for high margin MEMS from 2008-13, he said: “We don’t believe MEMS market is immune to the current market conditions. It went down 5 percent in 2008 and 8 percent in 2009. The last quarter of 2008 was very bad.

Not only automotive applications are suffering, the entire industry is suffering. Only a few segments in consumer electronics and cell phones, etc. are doing better. However, going 2010 onward, MEMS industry will see +11.5 percent growth from 2009-2013.

Coming to MEMS market by applications, mobile and CE are driving growth. They are not going down in 2009, but slightly growing by 4-5 percent. Most other applications are going down, however. Areas like military and civil aerospavce, medical electronics, industry and process control, are interesting.

High margin MEMS is growing as fast as CE and mobile MEMS — 16 percent CAGR from 2008-2013. Industry and process control will be biggest part of the market by 2013. Also, medical electronics and military and civil aerospace should do well, added Bouchaud.

This market should be more stable than what it is presently. MEMS wafer probe cards, e.g., is extremely dependent on the memory market, which has been going up and down. Demand for memory has been bad in 2008 and 2009.

He also presented a brief overview of the top suppliers for high margin MEMS — such as Formfactor, Honeywell, GE Sensing, FLIR Systems, Epson, DRS, JDSU, BAE Systems, MEAS and Panasonic.

Formfactor is the leading manufacturer, especially in wafer probes. Honeywell is a significant player, strong in accelerometers, gyroscopes for aerospace and defense applications. GE Sensing is a leader for pressure sensors for medical applications. FLIR Systems is a leader in microbolometers for industrial and defence application.

Epson is a leader for professional inkjet applications. Panasonic is a key player in pressure sensors, accelerometers, etc. JDSU is a purely fabless optical MEMS company. MEAS has a number of products — accelerometers, vibration sensors, flow sensors, etc.

Bouchaud also discussed the leading MEMS foundries for higher MEMS, where iSuppli calculated the revenue for high margin MEMS products only. These include companies such as Microalyne, Silex, IMT, Tronics, GE Sensing, Coilbrys, Menscap, Dalsa, Honeywell, Semefab, Microfab Bremen, X-Fab, MEAS and APM.

This is a very diverse picture, he said. “We have pure play foundries — purely focused on high end MEMS markets, such as Micralyne, IMT, Tronics.” Another group is covering both high end and volume markets such as consumer and automotive — like Silex, Dalsa, APM, etc. Then, there are those with mixed models, which are IDMs, such as GE Sensing, Colibrys, Memscap, Honeywell, etc.

High margin markets attract volume suppliers!
– Accelerometers/vibration sensors for health and structure monitoring, industry, medical, asset monitoring.

* This market was once dominated by expensive piezo-resistive MEMS sensors (like Endevco, MEAS).
* Volume accelerometer suppliers now try leveraging low power low cost and robust CE sensor experience into indl applications.
* VTI, ADI, Freescale, ST are now quite active.
* Accelerated price erosion is now happening due to the arrival of the big players. It could be a threat to established suppliers who have been having nice margins so far.

– High margin foundry work, e.g., for biomedical, industry, aerospace and defence, optical telecom.

* Traditionally, it was in the hands of Micralyne, Tronics, IMT.
* Silex historically on high margin markets, but has been moving to higher volume for customer during 2006 to 2008.
* Dalsa and Sony historically focused on higher volumes like consumer and automotive.
^ Raging competition on high volume MEMS with newcomers such as TSMC, UMC, SMIC.
* Medium sized players like Dalsa and Sony now increasing efforts to develop foundry biz with high margin MEMS due to the tough competition.

Snapshots of high margin markets
Bouchaud alsp provide snapshots on high margin MEMS markets.

The optical MEMS for telecom networks is said to be in transition. Poducts Includes VOAs, 1xN switches, NxN switch matrices, WSS ROADM, tunable filters, etc. There was collapse in Q4-08 through to Q2-09. However, there was an uptake in Q3-09. The key trend here is the increase in data rates from 10GB/s to 40GB/s to 100GB/s. This has had no impact on MEMS shipment. The only impact has been the specs — for some devices like VOAs.

Consolidation of market has been encouraged by system companies. Yet another impact is the arrival of Chinese suppliers who are driving price erosion on simple products. Overall, the market is flat for devices, with one exception being WSS ROADM — which will double in units. Finally, fabless models are dominating.

Biomedical applications not immune to economic crisis. They were given a lot of attention in the last decade. Bouchaud covered two groups:

* Microfluidics include silicon and glass microfluidics for drug discovery, POC diagnostics and drug delivery (inhalers, pumps, micro-needles).
* Sensors and switches include pressure sensors, accelerometers, thermopiles, flow sensors, switches and microbolometers.

As for trends, there was an inventory reduction in 2009 — up only 2pc in 2009. Also, the medical system companies have also been trying to cut costs. In microfluidics, drug discovery remains small. In POC diagnostics, there is no mass market with Si/glass. Also, ST’s chip hit the market. In drug delivery, there are over 5mn MEMS inhalers in 2009. As for micro pumps, the industry has currently adopted a ‘wait and see’ approach.

Pressure sensors are dominating in this segment. It is well established for blood pressure, sleep apnea monitoring, etc. There are strong R&D efforts for implantable wireless.

Next, inertial MEMS are gaining momentum over conventional sensors. Among the vibration/shock sensors, the piezoelectric (non MEMS) and piezo resistive still dominate. There has been slow adoption of silicon capacitive. Further, microfluids based accelerometers and gyroscopes are coming soon. As for the WSN (wireless sensor network) opportunity, iSuppli feels that the opportunity will remain modest over the next four years.

Some other trends include, geophones, which are dependant on oil price; MEMS accelerometers and gyroscopes growing in munitions guidance; MEMS gyro is displacing fiber optic gyro in defense and aeronautics; and pacemakers and defibrillators present the best opportunity in medical.

Supplier landscape has been shaken for microbolometers — yet another key trend. This is a double digit growth market — $300mn likely by 2013. Among applications, defense applications are dominating, while building inspection presents a good opportunity. However, automotive applications will remain niche.

Some other trends in this area include: export eased for US suppliers in 2008 and 2009 — good news for them, but bad news for Europen suppliers; standardization for firefighters; emergence of Chinese suppliers; prices remain high; limited adoption of WLP (wafer level packaging) still there.

Lastly, Bouchaud discussed MEMS switches, which are said to be in false start for high end applications. This segment did not really take from 2006-09. Today, it is mostly used for medical applications.

As a number of companies have either stopped production or gone bankrupt, the ATE and RF industry are now very skeptical. Also, ADF (automated distribution frameworks) opportunity is dead; and basically, companies in this area are now gone. ATE remains the best opportunity for high end applications. Defence applications will emerge from 2012 onward.