[time-nuts] New NIST Time Code to Boost Reception for Radio-Controlled Clocks

[Jim Lux]

On 5/31/14, 5:48 PM, Bob Camp wrote:
> Hi
>
> A thousand chips at $1 a chip is a very different thing than a thousand chips at $100 a chip. The next issue might be that they only have them in die form. The issue after that probably is that you really want the version 3 (or 9) chips that actually work with all the modulation schemes. I’ve been down the road with a number of similar chips that took *many* rev’s before they really did what they were intended to do. Many millions of dollars a pass times 3 or 6 passes is an whole different world …..
>
>

These days, though, the "per spin" cost is substantially lower, and the 
number of spins has been reduced, assuming you're forking out the many 
$M /year for the design tools.  I've been given to understand that an RF 
ASIC spin in CMOS (which works up to low microwave frequencies) is 
around $100k and would get you a wafer of dice.