You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
With 7nm is sight, the improvement in semis from shrink is coming to a close just on physical limitations. Advanced materials and 2D architectures that hold the promise of vastly increased speed will be the next frontier. But by far the largest new area will be in Mems and incorporating many of them directly in SOCs. Shrinking Mems will be a huge growth sector and I feel this is why Morris Chang said Mems would be one of the three largest opportunities for modern industry. The vast performance increases and price decreases that the semi sector has brought to the computational world will be brought to the physical world. We have just seen the very tip of this in display, solar, imaging, sound and a large variety of physical sensors. The scope and breadth of the Mems world while large now, is set to increase dramatically. Transportation of all forms and types have barely tapped the market that's yet to come. The human body will also be another mega market. The whole semi industry has been based on an unrelenting drive to lower cost and increase performance of everything it touches and with the combinations of Mems/Semi technology there is still an untapped market of several multiples of the current market as this drive extends further into the physical world.
This shift will create demand for new equipment, software and skill sets, while creating untold wealth. More than any other site I have read, the SemiWiki community has been more open minded and flexible and these factors combined with a new mind sets on education and training are the keys to the future.
The SemiWiki community has an unprecedented opportunity for continuing the ever accelerating growth in scope the semi sector has started and mastered as it is about applying accelerating knowledge at an ever accelerating rate. This speed of change and application to change the company, social, financial and even personal mind sets to change to fully realize the vast potential before us.
Agreed Arthur. If you consider digital computation a (highly) successful app for semiconductor engineering, then MEMS has the potential to be another comparably significant app. The digital app is constrained to processing what it is provided - it is a brain without senses (touch, smell taste, ...). The MEMS app can provide those senses in as many ways as we can imagine, and more.
Bernard, Thinking about it, I feel the Mems industry might even end up dwarfing the computational semi industry. Just giving materials the ability to change properties will in all probability become much larger than the computational semi sector. These will be a cross between Mems and active/passive phase change materials. Standard semis where just the first stage in giving materials active properties. Semis are just the first baby steps in advancing nanotechnology. Semi companies will have the edge, for they have a decades head start and the race is theirs to loose. Only the semi companies that have tunnel vision will overlook such a lucrative market, even if the license the technology to others. Active materials will be one of the hot new growth areas.
MEMS functions are done with predominantly mature process nodes like 180nm and for sensors (Inkjet print heads, accelerometers, gyroscopes, microphones, pressure sensors, displays, optical switching, bio, interferometric modular display, fluid acceleration, energy harvesting, ultrasound transducers). There is little economic drive to integrate MEMS onto leading-edge process nodes along with a billion digital gates. Look at the A-series of chips from Apple, the don't contain any MEMS because the MEMS features are in separate chips because of cost and performance.
Dan, I agree with your basic premise. Most of the sophisticated mems will be on larger nodes as will the computational and control parts for this will make the SOC of the future so cheap they will have wide spread and diverse markets. Labs of all types for everything from home use to commercial use will become wide spread as we have already seen in medical test equipment for professional and home use. It is already seeing wide spread use in auto diagnostics. Most devices in the future will have cheap built in diagnostic,s from large expensive vehicles to hand held and implanted devices. It's about cost and the large nodes are inexpensive which is why the market penetration will be very large. Phase change materials will be made using the same technologies used to make solar panels and displays. A roof that reflects and absorbs energy by different methods at different times may be the norm in the future at reasonable economic cost due to the unmatched ability of the semi sector to introduce new technologies, find new uses for mature technologies, improve performance and radically drive down costs relentlessly. I don't disagree with you at all on the mechanics, it's just we will see new business and product models.
Also we will see the sophistication of Mems take quantum leaps, just as we have seen in the computational world.
The only limits into the breadth and depth of new products will be the imagination and the Silicon Valley culture has proven almost limitless in this area. Some Mems will be in very large formats using the same technology we use to build solar panels and display screens. When sophisticated 3D printing is added to the equation, the number of future product classes shows a very bright future for almost all semi technologies and a great number of other fields that have been just waiting for new thinking and breakthroughs.
