> I'm sure there are a lot of transistors in their IC to handle all the phase > tracking and time decoding. They aren't doing phase tracking, just decoding time data. All that crap about Atomic Time is just advertising BS. Moore's Law applies to transistors. It doesn't matter if you use them for analog or digital. There is a wonderful concept: Pad Limited. Suppose your design takes N pads for bond wires to the outside world. Start with the logic for your design. Squish it down to a sensible shape. Now put the pads around it. Pads are big (relative to transistors). If they bump into each other, you may have empty room left over inside the pad ring. In a mature process, the yield is close to 100% and doesn't depend on how many transistors you put in that empty space so you might as well use it. If you don't have empty room today, wait a bit. Transistors are shrinking faster than pads. Watches are the classic example. They have lots of pads because each segment on the display needs its own wire. So you get watches that can keep track of days of the month and know about leap years. -- These are my opinions. I hate spam.

Hi > On Jul 30, 2020, at 5:11 PM, Hal Murray <hmurray@megapathdsl.net> wrote: > > >> I'm sure there are a lot of transistors in their IC to handle all the phase >> tracking and time decoding. > > They aren't doing phase tracking, just decoding time data. All that crap > about Atomic Time is just advertising BS. > > Moore's Law applies to transistors. It doesn't matter if you use them for > analog or digital. > > There is a wonderful concept: Pad Limited. How about “dicing limited” ? Turns out that your silicon wafer always needs to be chopped into pieces. That’s a messy mechanical process. It really hasn’t changed a lot over decades. A die that’s 1 mm on a side is about the smallest economical part. Go smaller and you “saw cut” simply is throwing away more silicon …. Bob > Suppose your design takes N pads > for bond wires to the outside world. Start with the logic for your design. > Squish it down to a sensible shape. Now put the pads around it. Pads are big > (relative to transistors). If they bump into each other, you may have empty > room left over inside the pad ring. In a mature process, the yield is close > to 100% and doesn't depend on how many transistors you put in that empty space > so you might as well use it. > > If you don't have empty room today, wait a bit. Transistors are shrinking > faster than pads. > > Watches are the classic example. They have lots of pads because each segment > on the display needs its own wire. So you get watches that can keep track of > days of the month and know about leap years. > > -- > These are my opinions. I hate spam. > > > > > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com > and follow the instructions there.