🚀 go-pugleaf

     Consider an OR gate each of whose inputs controls a radio
transmitter of frequency f.  The output of the OR gate is turned
on or off by a radio receiver tuned to the frequency f.  If one
(or more) of the inputs triggers a transmittion the output will
be turned on.  Any number of such OR gates can be built, each
with its own unique (private) frequency f, and each with an
arbitrary number of inputs.
     AND gates can be built from OR gates by adding inverters to
the inputs and output.  Again, fan in can be arbitrarily large
with no loss of speed.
     A complete computer can be assembled from AND terms OR'd
together.  Computer inputs (some of which may be inverted)
communicate with various AND gates by matched radio transmitters
and receivers, each set to its own frequency as needed.
     Similarly, connections can be made, from the outputs of various
AND gates to the inputs of OR gates, by matched transmitters and
receivers, each set to the proper frequency as needed.
     The "program" consists of the various matching transmitter and
receiver frequencies that couple inputs to ANDs and ANDs to ORs.
With an arbitrary number of frequencies available any number of
inputs and outputs are possible.
     It would seem that, neglecting transit time for the "radio
signals" (or light signals), any size computation could be accomplished
in constant time. Is this correct?  Can any computation be performed
in constant time?
     I realize, of course, that developing the program (i.e. tuning
the transmitter and receiver frequencies) can take a long time
indeed.  Feedback is also absent.










Finally,
connections are made fron the outputs of variou


Sent via Deja.com http://www.deja.com/
Before you buy.