5 ms·

Is the speed of light for light-in-some-substrate sufficiently faster than the speed of light for electrical charge? Is there any speed penalty for converting b

by mmt 8y ago

Is the speed of light for light-in-some-substrate sufficiently faster than the speed of light for electrical charge? Is there any speed penalty for converting between the two?

m1573rp34130dy 8y ago
the speed of light is always the speed of light, you have to observe at sufficiently small scale to make this observation. light in a medium travels between obstacles in a less than linear trajectory thus has an "apparent" average velocity. conversion from photonic to electronic can be very fast such as photonic emission during valence electron demotion, but in a practical case a junction has a latency and a fuzziness due to thermalnoise and quantum uncertainty...that is where the penalties will occur... conceptually there are 2 ends of a stick, does light actually move, or does light stand still while spacetime unfolds at "the speed of light"
- mmt 8y ago
  So, for a lay person, what's the summary? Does using optics as an interconnect in an otherwise electrical system add noticeable latency?
  - m1573rp34130dy 8y ago
    you are only as fast as your slowest segment... if you are moving single bits there is no problem... if you are using pulse width modulation or some other method of signal fidelity to stream packets then you will have to explore latency issues... if you are talking about single photons transverting to single electrons...you will have issue due to quantum interference... ...optocoupler latency is important and must be considered in your design, if you are working at circuit board level... im wondering just what your implementation is, are you talking "audiophile" equipment, are you talking digital data transmission... is it a human eye/ear noticing latency or is it an electronic component, that will consequently influence the smooth operation of the circuit?
    - mmt 8y ago
      I'm not the OC, but my understanding of the context is within an integrated circuit, possibly within a multi-die package, and as an alternative to smaller unit sizes. > is it an electronic component, that will consequently influence the smooth operation of the circuit? So, it would be this.
      - m1573rp34130dy 8y ago
        so it could be serial or parrallel data, typically serial is faster than parallel out of real world considerations its expensive to build a faster high quality serial link but it is cheaper to build a lower quality element for use in a high bandwidth parallel latch.. ...a parallel bus has settling time [~latency] that must be completed before latch values are ^valid^ and can/should be available to an input buffer for this reason you will have a maximum frequency threshold for statisically valid data...for parallel this is an average of each lane in the bus...for serial this is the overall latency of the junction operation... this is only considering the Tx>>Rx junctions...the signal path between must also be considered...potentially quantum chips would pump individual photons and tunnel individual electrons...or when we get there, we could talk about entanged photons Tx Rx bidirectionally through whatever it is that "tangles" between photonic superpositions...
        mmt 8y ago
        All this may be true, but I can't tell one way or the other, nor can I determine if it's actually applicable to the context of the discussion/article. It all seems very low-level. As a lay person (without EE or physics training), I'm still just looking for a high-level summary, as there's no way I'll learn anything useful, otherwise.
        m1573rp34130dy 8y ago
        there is a whitepaper typically issued for electronic components lets say a transistor and a latency, or frequency will be stated lets say 900MHz. that is a product of the over all switching latency of the transistor. so when current is applied to the gate lead of a transistor it takes a small amount of time for the gate to energize, another small amount of time for the gate portion of the transistor to propagate excitation to the semiconductor portion and allow conductivity to occur... ...it then takes time to tear it all down again...this is the latency of the transistor...in practical measure maximum frequency is stated so 900MHz means you can switch on then off again 900,000 times in 1 second. ...it gets technical again regarding nyquist frequency theory and weather you are Tx/Rx ing signal, or just generating noise by cramming changes in potential sardine style right next to each other so you cant tell one logic pulse from the previous or the next pulse... https://en.wikipedia.org/wiki/Transistor#Transistor_as_a_switch https://en.wikipedia.org/wiki/Transistor#Transistor_as_a_swi... ...latency matters when a chip has to operate so quickly that someportion of the chip is left waiting for another portion to finish its job... if you want information[signal] there is a required latency. if you flick a switch on, you have to wait for it to be flicked off before you can flick it on again...try this with a light switch, flick it on count to one flick it off count to one again flick switch on again...you should observe a light bulb light up 1sec, go dark 1sec, light up again..etc. ...now do this, flick the light switch a rapidly as possible up and down over and over again, if you do it fast enough the light will look very different it will seem to be illuminated or constantly flickering but not going dark...this is due to latency of the lightbulb filament, there is a frequency of switching that will break current and resend current to the light bulb faster than the light bulb can deenergize and darken...

by mmt 8y ago

Is the speed of light for light-in-some-substrate sufficiently faster than the speed of light for electrical charge? Is there any speed penalty for converting between the two?

m1573rp34130dy 8y ago

the speed of light is always the speed of light, you have to observe at sufficiently small scale to make this observation. light in a medium travels between obstacles in a less than linear trajectory thus has an "apparent" average velocity. conversion from photonic to electronic can be very fast such as photonic emission during valence electron demotion, but in a practical case a junction has a latency and a fuzziness due to thermalnoise and quantum uncertainty...that is where the penalties will occur... conceptually there are 2 ends of a stick, does light actually move, or does light stand still while spacetime unfolds at "the speed of light"

mmt 8y ago
So, for a lay person, what's the summary? Does using optics as an interconnect in an otherwise electrical system add noticeable latency?
- m1573rp34130dy 8y ago
  you are only as fast as your slowest segment... if you are moving single bits there is no problem... if you are using pulse width modulation or some other method of signal fidelity to stream packets then you will have to explore latency issues... if you are talking about single photons transverting to single electrons...you will have issue due to quantum interference... ...optocoupler latency is important and must be considered in your design, if you are working at circuit board level... im wondering just what your implementation is, are you talking "audiophile" equipment, are you talking digital data transmission... is it a human eye/ear noticing latency or is it an electronic component, that will consequently influence the smooth operation of the circuit?
  - mmt 8y ago
    I'm not the OC, but my understanding of the context is within an integrated circuit, possibly within a multi-die package, and as an alternative to smaller unit sizes. > is it an electronic component, that will consequently influence the smooth operation of the circuit? So, it would be this.
    - m1573rp34130dy 8y ago
      so it could be serial or parrallel data, typically serial is faster than parallel out of real world considerations its expensive to build a faster high quality serial link but it is cheaper to build a lower quality element for use in a high bandwidth parallel latch.. ...a parallel bus has settling time [~latency] that must be completed before latch values are ^valid^ and can/should be available to an input buffer for this reason you will have a maximum frequency threshold for statisically valid data...for parallel this is an average of each lane in the bus...for serial this is the overall latency of the junction operation... this is only considering the Tx>>Rx junctions...the signal path between must also be considered...potentially quantum chips would pump individual photons and tunnel individual electrons...or when we get there, we could talk about entanged photons Tx Rx bidirectionally through whatever it is that "tangles" between photonic superpositions...
      - mmt 8y ago
        All this may be true, but I can't tell one way or the other, nor can I determine if it's actually applicable to the context of the discussion/article. It all seems very low-level. As a lay person (without EE or physics training), I'm still just looking for a high-level summary, as there's no way I'll learn anything useful, otherwise.
        m1573rp34130dy 8y ago
        there is a whitepaper typically issued for electronic components lets say a transistor and a latency, or frequency will be stated lets say 900MHz. that is a product of the over all switching latency of the transistor. so when current is applied to the gate lead of a transistor it takes a small amount of time for the gate to energize, another small amount of time for the gate portion of the transistor to propagate excitation to the semiconductor portion and allow conductivity to occur... ...it then takes time to tear it all down again...this is the latency of the transistor...in practical measure maximum frequency is stated so 900MHz means you can switch on then off again 900,000 times in 1 second. ...it gets technical again regarding nyquist frequency theory and weather you are Tx/Rx ing signal, or just generating noise by cramming changes in potential sardine style right next to each other so you cant tell one logic pulse from the previous or the next pulse... https://en.wikipedia.org/wiki/Transistor#Transistor_as_a_switch https://en.wikipedia.org/wiki/Transistor#Transistor_as_a_swi... ...latency matters when a chip has to operate so quickly that someportion of the chip is left waiting for another portion to finish its job... if you want information[signal] there is a required latency. if you flick a switch on, you have to wait for it to be flicked off before you can flick it on again...try this with a light switch, flick it on count to one flick it off count to one again flick switch on again...you should observe a light bulb light up 1sec, go dark 1sec, light up again..etc. ...now do this, flick the light switch a rapidly as possible up and down over and over again, if you do it fast enough the light will look very different it will seem to be illuminated or constantly flickering but not going dark...this is due to latency of the lightbulb filament, there is a frequency of switching that will break current and resend current to the light bulb faster than the light bulb can deenergize and darken...