Photonic Equivalent of "Leakage Current"?

[Radium]

With a billion CPUs, theleakage currentwould kill you. If you want
real processing speed at low power, you should look at using 3 phase
clocks. There are several advantages to this. You only have to swap
two lines of the 3 phase clock to invert the order. This means that
the processor can back step. It doesn't really make a general purpose
computer but it would be very handy if you were playing jeopardy.

Just out of curiosity, lets say [hypothetically], I had a PC that used
laseronics [photonics using lasers and without any LEDs] in place of
electronics. If this laseronic computer uses "parallel-Hz" would it
run into something similar to "leakage current"? If so, what is the
optical equivalent of leakage current?


Thanx,

Radium

 

[contrex]

With a billion CPUs, theleakage currentwould kill you. If you want
real processing speed at low power, you should look at using 3 phase
clocks. There are several advantages to this. You only have to swap
two lines of the 3 phase clock to invert the order. This means that
the processor can back step. It doesn't really make a general purpose
computer but it would be very handy if you were playing jeopardy.

Just out of curiosity, lets say [hypothetically], I had a PC that used
laseronics [photonics using lasers and without any LEDs] in place of
electronics. If this laseronic computer uses "parallel-Hz" would it
run into something similar to "leakage current"? If so, what is the
optical equivalent of leakage current?

Thanx,

Radium

Radium, for whom killfiles were invented!
<plonk>

 

[Michael A. Terrell]

Radium, for whom killfiles were invented!
<plonk>

ADT


--
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

 

[Rich Grise]

Just out of curiosity, lets say [hypothetically], I had a PC that used
laseronics [photonics using lasers and without any LEDs] in place of
electronics. If this laseronic computer uses "parallel-Hz" would it
run into something similar to "leakage current"? If so, what is the
optical equivalent of leakage current?

Nobody knows. You'll have to build one and find out!

Please report back and let us know the results.

Good Luck!
Rich

 

[Bernd Paysan]

Just out of curiosity, lets say [hypothetically], I had a PC that used
laseronics [photonics using lasers and without any LEDs] in place of
electronics. If this laseronic computer uses "parallel-Hz" would it
run into something similar to "leakage current"? If so, what is the
optical equivalent of leakage current?

Optical damping. Well, leakage current is unavoidable when you can't turn
supplies off, because you need it to remember states of flip-flops or such
like. If you figure out how to have a persistent storage within a
flip-flop, that doesn't need power, you can avoid leakage by powering up
the systen only when you need it.

 

[MooseFET]

With a billion CPUs, theleakage currentwould kill you. If you want
real processing speed at low power, you should look at using 3 phase
clocks. There are several advantages to this. You only have to swap
two lines of the 3 phase clock to invert the order. This means that
the processor can back step. It doesn't really make a general purpose
computer but it would be very handy if you were playing jeopardy.

Just out of curiosity, lets say [hypothetically], I had a PC that used
laseronics [photonics using lasers and without any LEDs] in place of
electronics. If this laseronic computer uses "parallel-Hz" would it
run into something similar to "leakage current"? If so, what is the
optical equivalent of leakage current?

That would be "leakage light".

Since a photon must contain energy and the usual Plank's constant
applies to it, you can never have your 1Hz machine. You get a 10^15
Hz machine when you start using light as your clock source. Current
semiconductor technology is less decades below 10^15Hz than it is
above 1Hz so you need to consider if it will be lasers or transistors
that will make the light.

Using transistors to make the light can have some advantages. If you
wish to experiment with such ideas, I suggest you scale down in
frequency and up in the spacial dimensions so that you can use
macroscopic parts like a 2N2222 and lengths of wire to make the
radiators. Ironoxide will make your nonlinear magneto-optical
effects.

 

[krw]

Hi:

Below is an example of "parallel Hz"

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.
With a billion CPUs, theleakage currentwould kill you. If you want
real processing speed at low power, you should look at using 3 phase
clocks. There are several advantages to this. You only have to swap
two lines of the 3 phase clock to invert the order. This means that
the processor can back step. It doesn't really make a general purpose
computer but it would be very handy if you were playing jeopardy.

Just out of curiosity, lets say [hypothetically], I had a PC that used
laseronics [photonics using lasers and without any LEDs] in place of
electronics. If this laseronic computer uses "parallel-Hz" would it
run into something similar to "leakage current"? If so, what is the
optical equivalent of leakage current?

That would be "leakage light".

Since a photon must contain energy and the usual Plank's constant
applies to it, you can never have your 1Hz machine. You get a 10^15
Hz machine when you start using light as your clock source. Current
semiconductor technology is less decades below 10^15Hz than it is
above 1Hz so you need to consider if it will be lasers or transistors
that will make the light.

Why would a clock necessarily be at the frequency of the light? Is
the data rate in fiber optics at the frequency (or some integer
fraction of the frequency) of the light?

 

[MooseFET]

(e-mail address removed) says... [....]

Since a photon must contain energy and the usual Plank's constant
applies to it, you can never have your 1Hz machine. You get a 10^15
Hz machine when you start using light as your clock source. Current
semiconductor technology is less decades below 10^15Hz than it is
above 1Hz so you need to consider if it will be lasers or transistors
that will make the light.

Why would a clock necessarily be at the frequency of the light?

Remember this is some dream machine we are talking about here. If you
don't have the clock speed equal to the frequency of the light, you
are wasting all those edges. Skipping edges on the "clock" doesn't
count as meeting Radiums goals.

Is
the data rate in fiber optics at the frequency (or some integer
fraction of the frequency) of the light?

In nearly every case, a bit on a fiber is carried by many photons[1].
The exact number of photons varies from bit to bit and the photons are
not all at exactly the same wavelength. Basically, we have a
fractional divider situation. Some bits have a different number of
cycles than others.

[1] The exceptions are out of the scope of this discussion.

 

[krw]

(e-mail address removed) says... [....]

Since a photon must contain energy and the usual Plank's constant
applies to it, you can never have your 1Hz machine. You get a 10^15
Hz machine when you start using light as your clock source. Current
semiconductor technology is less decades below 10^15Hz than it is
above 1Hz so you need to consider if it will be lasers or transistors
that will make the light.

Why would a clock necessarily be at the frequency of the light?

Remember this is some dream machine we are talking about here. If you
don't have the clock speed equal to the frequency of the light, you
are wasting all those edges. Skipping edges on the "clock" doesn't
count as meeting Radiums goals.

I didn't see where he wanted to use every "edge" of the light wave.
Did I miss that?

Is
the data rate in fiber optics at the frequency (or some integer
fraction of the frequency) of the light?

In nearly every case, a bit on a fiber is carried by many photons[1].
The exact number of photons varies from bit to bit and the photons are
not all at exactly the same wavelength.

The laser makes them pretty close. Ok, there is some phase noise in
there. So?

Basically, we have a
fractional divider situation. Some bits have a different number of
cycles than others.

[1] The exceptions are out of the scope of this discussion.

Nope. This is the Usenet. There is no "scope of the discussion".
;-)

Damn! They're shooting 105s down the road from here and the cat is
wondering what's going on.

 

[MooseFET]

(e-mail address removed) says... [....]
Since a photon must contain energy and the usual Plank's constant
applies to it, you can never have your 1Hz machine. You get a 10^15
Hz machine when you start using light as your clock source. Current
semiconductor technology is less decades below 10^15Hz than it is
above 1Hz so you need to consider if it will be lasers or transistors
that will make the light.
Why would a clock necessarily be at the frequency of the light?

Remember this is some dream machine we are talking about here. If you
don't have the clock speed equal to the frequency of the light, you
are wasting all those edges. Skipping edges on the "clock" doesn't
count as meeting Radiums goals.

I didn't see where he wanted to use every "edge" of the light wave.
Did I miss that?

In nearly every case, a bit on a fiber is carried by many photons[1].
The exact number of photons varies from bit to bit and the photons are
not all at exactly the same wavelength.

The laser makes them pretty close. Ok, there is some phase noise in
there. So?

You were asking about the number of cycles being an integer or not.
Since the frequency can't be said to be a constant, it is hard to say
anything about integer numbers of cycles. The bandwidth of the output
of a single mode VCSEL is many MHz wide. A multimode one has a
bandwidth in the GHz.

Number of photons <> different number of cycles

Agreed but in terms of both photons and cycles, different bits have
different numbers of them.

[1] The exceptions are out of the scope of this discussion.

Nope. This is the Usenet. There is no "scope of the discussion".
;-)

Ok, so you out smartass remarked me this time. Just wait for next
time.

 

[Radium]

:

Just out of curiosity, lets say [hypothetically], I had a PC that used
laseronics [photonics using lasers and without any LEDs] in place of
electronics. If this laseronic computer uses "parallel-Hz" would it
run into something similar to "leakage current"? If so, what is the
optical equivalent of leakage current?

That would be "leakage light".

Since a photon must contain energy and the usual Plank's constant
applies to it, you can never have your 1Hz machine. You get a 10^15
Hz machine when you start using light as your clock source. Current
semiconductor technology is less decades below 10^15Hz than it is
above 1Hz so you need to consider if it will be lasers or transistors
that will make the light.

Using transistors to make the light can have some advantages. If you
wish to experiment with such ideas, I suggest you scale down in
frequency and up in the spacial dimensions so that you can use
macroscopic parts like a 2N2222 and lengths of wire to make the
radiators. Ironoxide will make your nonlinear magneto-optical
effects.

Would "leakage light" cause damage to a photonic system in a similar
manner in which leakage current damages electronic systems.

Also, I doubt that the clock rate of a photonic system has to be that
of light. In clock rate, the "Hz" is the number of pulses per second

http://www.tiscali.co.uk/reference/dictionaries/computers/data/m0034449.html

"Clock rates are measured in megahertz (MHz), or millions of pulses a
second."

The clock rate of a CPU is usually determined by the frequency of an
oscillator crystal.

So if the crystal gives out on pulse per second, the clock rate is 1
Hz.

I could be wrong though.

 

[MooseFET]

Would "leakage light" cause damage to a photonic system in a similar
manner in which leakage current damages electronic systems.

Normally electrical leakage doesn't do any harm. Leakage of light
wouldn't either.

Also, I doubt that the clock rate of a photonic system has to be that
of light. In clock rate, the "Hz" is the number of pulses per second

Perhaps I didn't understand how you were defining things. If the goal
is the greatest posible speed for the lowest power, you need to
consider the energy that is in each photon. This would depend on the
wavelength.

http://www.tiscali.co.uk/reference/dictionaries/computers/data/m00344...

"Clock rates are measured in megahertz (MHz), or millions of pulses a
second."

The clock rate of a CPU is usually determined by the frequency of an
oscillator crystal.

So if the crystal gives out on pulse per second, the clock rate is 1
Hz.

In actual fact the frequency of the processor is rarely equal to the
frequency of the crystal. A crystal may be running at 20MHz and the
processor at 1GHz. The oscillator that runs the processor is kept on
frequency by referencing it to the 20MHz.

 

[Radium]

Normally electrical leakage doesn't do any harm. Leakage of light
wouldn't either.

Doesn't the leakage current cause the system to overheat?

Perhaps I didn't understand how you were defining things. If the goal
is the greatest posible speed for the lowest power, you need to
consider the energy that is in each photon. This would depend on the
wavelength.

400 nm wavelength

In actual fact the frequency of the processor is rarely equal to the
frequency of the crystal. A crystal may be running at 20MHz and the
processor at 1GHz. The oscillator that runs the processor is kept on
frequency by referencing it to the 20MHz.

So the clock rate and the rate at which the crystal oscillates don't
match each other? Which frequency is usually higher? What about the
frequency of one determines the frequency of the other?

 

[The little lost angel]

So the clock rate and the rate at which the crystal oscillates don't
match each other? Which frequency is usually higher? What about the
frequency of one determines the frequency of the other?

The effective clock rate is usually higher because it is derived as a
multiple of the basic crystal oscillation frequency. i.e. the crystal
is running at 8.33Mhz but the system bus could be clocked at 24x at
about 200Mhz.

 

[Bob Myers]

Doesn't the leakage current cause the system to overheat?

Absolutely. This is why every electronic device you've
ever seen turns into a puddle of molten metal and silicon
within seconds of being turned on.

400 nm wavelength

But what if I don't like that color?

So the clock rate and the rate at which the crystal oscillates don't
match each other? Which frequency is usually higher? What about the
frequency of one determines the frequency of the other?

And why not? He's earned every penny of it!

Bob M.

 

[Rev. 11D Meow!]

3

the answer is always 3.

With a billion CPUs, theleakage currentwould kill you. If you want
real processing speed at low power, you should look at using 3 phase
clocks. There are several advantages to this. You only have to swap
two lines of the 3 phase clock to invert the order. This means that
the processor can back step. It doesn't really make a general purpose
computer but it would be very handy if you were playing jeopardy.

Just out of curiosity, lets say [hypothetically], I had a PC that used
laseronics [photonics using lasers and without any LEDs] in place of
electronics. If this laseronic computer uses "parallel-Hz" would it
run into something similar to "leakage current"? If so, what is the
optical equivalent of leakage current?


Thanx,

Radium

 

[Rev. 11D Meow!]

The answer is still 3 .

Doesn't the leakage current cause the system to overheat?


400 nm wavelength


So the clock rate and the rate at which the crystal oscillates don't
match each other? Which frequency is usually higher? What about the
frequency of one determines the frequency of the other?