As some have noted, CMOS oscillators are already in the market from IDT. In fact, CMOS oscillators have replaced XTALs in nearly all USB 2.0 PHYs and the expectation is that the same will be true for USB 3.0. As IDT announced a few years ago, it licensed its CMOS oscillator technology to several USB PHY developers while also supplying die products to others. The point here is that CMOS oscillators are well-suited to wireline communications. As others have noted, they are not well-suited to wireless applications due to the need for low close-to-carrier phase noise. Thus, the change in strategy from eoSemi comes as no surprise. Nevertheless, CMOS oscillators are pervasive, though I have never seen a CMOS oscillator product from eoSemi. To my knowledge, only IDT and its licensees have been shipping CMOS oscillators and products with embedded CMOS oscillators in volume.
A little more detail on MEMS-and-Crystal free CMOS oscillators:
There is admittedly little chance that such oscillators will displace crystals in situations where sideband noise is critical. The target for eoSemi is timing applications where SXOs or other off-chip components would previously have been used, particularly where it is desirable to integrate the timing circuitry with other parts of the system.
eoSemi has already demonstrated performance that is superior to anything we know of from IDT. This is a consequence of a range of proprietary techniques, including circuit designs that minimise the effects of variations in CMOS transistor performance, and specialised strain-compensation algorithms. We now look both to further improve the performance and to minimise production overhead.
Blog Doing Math in FPGAs Tom Burke 23 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 ...