golfdog - I invited you to download SiTime's performance reports (http://www.sitime.com/support/performance-measurement-report) to see for yourself how our jitter, frequency stability, aging and STFS (Allan Deviation) performance meets the requirements of all computing, consumer, enterprise, storage, and Ethernet applications. You'll have to register to get access to this data.
Electronics companies are also converting to MEMS timing because of another reason - Resilience. See here (http://www.sitime.com/company/technology-overview/benefits/40x-more-resilient) for measured data on how SiTime's Silicon MEMS timing solutions are up to 40 times more superior than quartz devices.
The timing market (resonators, oscillators, clock generators, buffers) is already $6B in size. Customers are converting rapidly from legacy quartz solutions to advanced MEMS solutions because they benefit in performance, resilience, reliability and cost. This is driving the rapid growth of the Silicon MEMS timing industry.
Exec VP, Marketing, SiTime Corp.
MEMS appear to be well suited for high-vibration environments, to non-critically timed applications and applications where signal-to-noise ratios are not important.
Applications that have complex modulation schemes, very high-speed communication, or that require excellent signal-to-noise performance will continue to be clocked by crystal based oscillator devices that have exceptional high Q and excellent temperature stability characteristics found in Quartz.
The two technologies are not interchangeable. Frequency variations from both the lower Q and digital temperature compensation in the MEMS oscillators result in frequency fluctuations that are unacceptable in most applications.
(Would require a form of digital adjustments in the PLL circuit to correct the frequency…)
Jitter is an important characteristic in most of today’s computing and RF applications and quartz based oscillators have jitter performance that is equal or exceeds the best oscilloscopes. MEMS will never meet this requirement.
Is there merit to MEMS in low end applications?…Yes there is….but they are limited. Just because this is a silicon device does not mean its inherent characteristics will eliminate the need for quartz based devices.
In 2010…the global market created $165 Million in revenues for MEMS…with the majority coming from the consumer and the automotive sensor application space.
I disagree with Karen that this will be a $20B market in 5 years. The Crystal based global timing market now is only $29B. The cost of MEMS at this juncture cannot keep pace with the lower cost of the new innovative quartz fabrication technologies. With the technology advantages of quartz crystal based devices in this ever evoloving wireless space we live in...the applications for MEMS are limited as noted.
I beg to differ - the reason there is specialization is because there is so much opportunity for the adoption of MEMS! It's not fragmented; it's growing and maturing and developing into a robust industry - predicted to top $20B by 2017! Yes there will be some consolidation but there will more partnership (see this article by Peter Clarke as an example) which will spur more opportunity for innovation and yes, more adoptions of MEMS.
Blog Doing Math in FPGAs Tom Burke 13 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...