My "Parallel Hz" Fantasy

[Radium]

Hi:

Below is an example of my "parallel hz"

http://img56.imageshack.us/img56/2427/clocksignalexample8is.gif

If each clock signal is 1 hz, and you have a billion of them, staggered
such that every 1ns part of the CPU can start, and finish, an
instruction - making the effective 'clock rate' 1 Ghz.

There is no real benefit to using a billion 1 hz clocks to make a clock
rate of 1Ghz and setting up such a system would be a large mess to CPU
designers. However, I find it fun to think about. Fun. Like a hobby, an
interest, etc.

In a "parallel Hz" device the bits maybe completely in serial and the
algorithms and tasks maybe totally non-parallelizable. However, the
frequency is still parallel.

The device I am proposing is completely serial except for the clock
rate.

My proposed device is completely serial except for the frequency. It
uses "parallel hz" but in terms of everything other than frequency, it
is totally serial and non-parallel. Only the clock rate is parallel.

Parallel hz = a method using N number of 1 hz clocks to gain a clock
rate of N hz.

My design has has a clock rate of 4 Ghz that is obtained by using 4
billion 1 hz
clocks. But otherwise, it is completely serial.

I recently crossposted my question to some other groups. One of the
responses was that my device would "leak" 3999999999/4000000000ths of
the time. What does "leaking" mean? What causes "leaking"?


Thanks,

Radium

 

[Bob Myers]

If each clock signal is 1 hz, and you have a billion of them, staggered
such that every 1ns part of the CPU can start, and finish, an
instruction - making the effective 'clock rate' 1 Ghz.

And how, pray tell, do you accurately "stagger"
a billion pulses 1 ns apart without have a 1 GHz clock
or better in the first place?

Bob M.

 

[Michael A. Terrell]

And how, pray tell, do you accurately "stagger"
a billion pulses 1 ns apart without have a 1 GHz clock
or better in the first place?

Bob M.

Well, he did say that it was a fantasy.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[Radium]

And how, pray tell, do you accurately "stagger"
a billion pulses 1 ns apart without have a 1 GHz clock
or better in the first place?

I wish I knew. Really. I thought that website described it in imagery.

 

[G. Schindler]

I wish I knew. Really. I thought that website described it in imagery.

Actually, the 1 nS stagger is the easy part as it could be implemented
utilizing the propagation time between elements..... for that matter you
could do the same thing with only one clock recombining multiple delayed
copies of the output signal.

The delay could be implemented by short lengths of coax or just PCB traces.

Not very practical, but a fun mind game.

 

[Radium]

Actually, the 1 nS stagger is the easy part as it could be implemented
utilizing the propagation time between elements..... for that matter you
could do the same thing with only one clock recombining multiple delayed
copies of the output signal.

The delay could be implemented by short lengths of coax or just PCB traces.

Not very practical, but a fun mind game.

So, IOW, it is physically-possible to have a 4Ghz clock rate obtain
solely by using 4 billion 1 hz clocks w/out there being an actual 4 Ghz
clock?

 

[Bob Myers]

So, IOW, it is physically-possible to have a 4Ghz clock rate obtain
solely by using 4 billion 1 hz clocks w/out there being an actual 4 Ghz
clock?

It is only barely within the realm of possibility; in practice,
getting everything to work out just right would make it a
nightmare. And there's really no point to it at all - it would
simply demonstrate that all you care about is a billion edges
per second, all parked at the right place - but so what?

Bob M.

 

[Radium]

It is only barely within the realm of possibility; in practice,
getting everything to work out just right would make it a
nightmare. And there's really no point to it at all - it would
simply demonstrate that all you care about is a billion edges
per second, all parked at the right place - but so what?

I am just in it for the fun.

 

[G. Schindler]

So, IOW, it is physically-possible to have a 4Ghz clock rate obtain
solely by using 4 billion 1 hz clocks w/out there being an actual 4 Ghz
clock?

Obviously, having to construct a billion of anything (except hamburgers)
would prove pretty daunting. On the other hand, if you wanted to
construct (for instance) a 4Mhz source from a 1Mhz source this technique
could be used easily since bulding four copies of the original would be
entirely practical.

Again, the real way to do this is to use one source and just have the
output go through multiple delay paths before being recombined.

Remember, I didn't say it was possible to do the 1GHz clock from the 1Hz
clock .... I only said that the 1nS stagger was the easy part. This
technique, however, would allow you to construct real circuits if the
numbers were kept a little closer to reality.

 

[Bob Myers]

I am just in it for the fun.

Sorry, guess I'm missing seeing where the fun comes
in in all this. A billion edges per second IS a 1 GHz
clock; your approach differs only in that it has the
distinct disadvantage of requiring a billions separate
wires to go along with those edges.


Bob M.

 

[Mark L. Fergerson]

So, IOW, it is physically-possible to have a 4Ghz clock rate obtain
solely by using 4 billion 1 hz clocks w/out there being an actual 4 Ghz
clock?

It's being done, but sideways, sort of:

http://www.multigig.com/pub.html

One clock, as many outputs as you want (/have room for), with
whatever phase delay you can arrange for depending where you pick them
off. Compare signals from different pickoff points; effectively a much
faster clock rate than the clock puts out.


Mark L. Fergerson

 

[[email protected]]

Obviously, having to construct a billion of anything (except hamburgers)
would prove pretty daunting.

Would prove pretty expensive too.