how do fiber cables transfer so much data at such speeds
they are tiny and fragile
 Shopping Cart Returner Shirt $21.68 |
 Nothing Ever Happens Shirt $21.68 |
Shopping Cart Returner Shirt $21.68 |
how do fiber cables transfer so much data at such speeds. they are tiny and fragile
Really fast pulses of light
They use light
because of the thightness the particles bounce less inside so they travel faster
they go to where the warm is
optical processing allows for more domains of information than radio frequency for one since you can also encode data in the polarization of light, radio heterodyning and signal processing is not there yet (ex:docsis)
light is an also extremely high frequency medium. If you try to modulate data on a 5ghz baseband over a 1ghz radio you physically cannot do it, you cannot change the voltages faster than 1 billion times per second peak to peak.
Light in fiber optic communications is on the order of 100 THz, given sensitive and fast enough transcievers the upper limit for a channel is extremely high and wide.
These days docsis is able to reach 1gbit similarly to most fiber networks thanks to improved modulation schemes and channel bonding. The return path is still shitty like it has always been though.
Fiber optics reach speeds in excess of 800GbE for single transcievers, 1.2T and 1.6T is on the horizon as soon as the IEEE and OIF get around to unfricking the super high baud rate SerDes. Optical add drop mux's will also carry an insane amout of data over one fiber pair. Some support 96x1.2Tbps channels some will do 128 (64 per band).
(G/X)PON is only comparable to docsis since it is one passive, direct detect non coherant "grey" optics and meant to provide wideband medium access on one fiber pair rather than gigantic point to point links like with carrier ethernet fiber links. Both are common in that it's ~1ghz modulation over an em medium but the key difference is between the radio and optical domains that let you do different things, and also run into their own problems like non linear effects for coherant optics.
*PON doesn't have a "fiber pair" it's a single fiber
Aren't the optics bit on the limit now and some OEM's are toying around co-packaged optics?
To a point yes. It's less so the optics more the bottleneck in SerDes and the on chip beachfront required to get data off the chips. Single organic chips are now doing >50Tbps and moving that much on and off chip is getting to be very hard.
Copackaging silicon photonics is the future for sure, on market high end transcievers have been copackaged optics for a while now anyways(but the link between the transcievers and forwarding elements is still electrical). Broadcom just came out with a fully "optical" switch which packages all the optics on the forwarding die.
But the drive for copackaging isn't actually an issue with the optics. Cmos optics are already and have been used for a while for these optical lasers like I said. It's that the electrical interconnects between chips is becoming a huge bottleneck. PAM4 has helped a lot but memory has started to become a shitshow, you need an interposer and then the classic climb to needing retimers becomes a question. Copackaging optics for chip to chip and chiplet interconnect is where most of the research and advances are from what I have seen.
>light is an also extremely high frequency
This is the real answer.
Ultimately the amount of data you can transfer per second depends on the bandwidth, i.e.: the difference between the highest and lowest frequency.
(See: Heisenberg's uncertainty principle).
Using light you easily get bandwidths in terahertz - far more than you can achieve using radio waves or electronics.
>Using light you easily get bandwidths in terahertz
Not for a very long time maybe 50+ years from now. The bandwidths are still comparable to radio mediums, like how radio has120mhz wide channels at 5Ghz basebands optical bandwidths aren't much larger and the baseband frequencies are only ever ~30Ghz some are higher but it's a huge challenge to get circuits that do super wide band high baud rate super high frequency dsp. It's just that the optical medium is currently a well of unlimited potential since there is no realistic upper bound for how wide and high frequency you can make channels. As long as you can makw a faster circuit you can keep using light unlike radio where eventually you hit a wall around 50-100Ghz.
The main distinction for datarate currently for comparable channel widths and baseband frequencies really is the superior modulation of light. It handles spread spectrum techniques better, polarization is a huge deal and the decision boundaries are better for QAM
>Using light you easily get bandwidths in terahertz
no, and it also depends on the modulation used, Black person.
wrong
bandwidth doesn't depend on modulation. It's just the width of the channel that you are sampling a signal from.
It's the same as IrDa only the light is traveling through the cable instead of through the air.
The brings up my constant posting of IrDa on the site. Why limit your light by binding it in a cable? Set it free and watch your data transfers travel even faster than fiber.
So one day we won't even need any cables?
it fast
Black body radiation… uuhh maxwells equations in vacuum or something
they're made of hedgehog
They're made of mirror tubes.
It's alien technology.
You didn't hear this from me.
Don't ask questions, citizen.
light travels at the speed of light, but electricity only travels at the speed of electricity
And wi-fi follows a ballistic trajectory
wi fi follows the shape of a weibull distribution
nah, pretty sure it’s ballistic
The individual electrons in moving current travel on the order of centimetres per second (drift velocity), but the signal itself travels, approximately at the speed of light.
Think of a human stampede; each person moves slowly but the signal of 'run' travels quickly through the crowd.
in fact, coax propagates signals at about 0.8c, whilst fibre only does about 0.6c. Fibre comes on top because that massive bandwidth, as
pointed out, which lets you stuff more data in per second. Fibre is like a slow truck, coax is like a fast motorcycle. Coax delivers the start of the data first, but in the long run, the fibre finishes the job first.
The speed of signal propagation in copper wire (~ 200,000 km/h) is not approximately the speed of light in a vacuum (~1,000,000,000 km/h).
The speed of light in a vacuum is irrelevant here, since fiber optic cables are not made of a fricking vacuum. They're glass, so you need to consider the speed of light in glass, which is a lot closer to the speed of an electrical pulse through copper.
>gets assblasted that "speed of light" without any qualifier is assumed to mean C
the area between your ears is made of a fricking vacuum
This board is about technology. The concept of speed of light changing in different materials is taught in highschool science class. If you don't even have that base level of knowledge what the frick are you doing here?
lol, no, I clearly understand that concept. You're just moving the goal posts, you stupid homosexual cum brian.
you lost that shit badly L
https://en.wikipedia.org/wiki/Velocity_factor
We're discussing optical fiber cable (0.67c) and coax cable (0.82c - 0.93c, depending on the type of coax).
had reasonable numbers.
photons are massless
that is not the scientific consensus
What? Yes it is. It's basically unanimous.
There is no such thing as a photon
Light is a wave of aether
It's like sound waves but at a metaphysical level
The transition of light from on to off is very abrupt and noise free
Your noise floor in general is very good because it's not possible for outside interference to leak in
direct detect systems for both optics and radios exist but coherant optics always trump direct detect. With direct detect you are limited to a few forms of digital mode keying at best, or just on off keying, where with coherant optics you have the full form of analogue modulation available which lets you encode a ton of data for low baud rates which is one of the main barriers for communication speeds.
so what's the state of optical computing?
nothing
there is a little fella inside of them who runs very fast carrying data packets back and forth. do not try to observe the little fella, though, the CIA will kill you. let him do his job in peace
The medium for which light travels is the answer. Glass is 1/3 as dense as copper. Thus allowing more bandwidth.
That's not even yet getting into single and multimode optic cables which makes an even Bigger difference
Fiber optic cables use light signals instead of electrical currents.
They consist of thin glass strands (optical fibers).
Light travels through total internal reflection.
Benefits include high speed, long distances, and immunity to electromagnetic interference.
Despite their delicate appearance, they are robust and essential for data transmission.
yes i used gpt, frick off
GPT got it wrong. It's a lower speed than electrical currents.
See
and
when did gpt say "faster speed", it said high speed, but nothing about it being faster?
It’s kind of worrying that coders and other people that should know how this stuff works essentially regard electricity and electronics as black magic
I don't think it's that concerning; these details about the physical stuff happening are highly abstracted away. Though perhaps it would still be useful to know about latency vs marshalling delay
I feel as if most CS programs should cover at least some basic circuits & electrodynamics
>electricity and electronics as black magic
I'm yet to hear an argument as to why it is not black magic
You are using light to send e-girl porn around the world
That is black magic
I see I see
So do we bring back witch trials?
People sending e-girl porn around the world with black magic are already subject to witch trials.
Any sufficiently advanced technology is indistinguishable from magic.
It's magic.
t. electronics engineer
The phlogiston in you cable cooks the packages
Well OP purple gets to send data,green gets to send data and so on.... Different wave lengths for one thing.Infinite amounts of colors and shades in-between. Lol
Light is pretty fast
Some nerds even say it's the fastest thing in the universe.
light go fast