can any of you answer this?

[z.monkey]

Howdy,

This was almost two years ago.

http://www.news.uiuc.edu/NEWS/06/1211transistor.html

So by now they are probably pushing 5 terahertz.  A terabit router is not the same thing as a single component clocking at 1 terahertz.  The router is going to use parallel channels to make that bandwidth.  The ethernet cables that the data is moving on are clocking at 250 megahertz for a 1 gigabit ethernet, 2.5 gigahertz for a 10 gigabit ethernet.  A terabit router would have a 1000, 250 megahertz channels or 100, 2.5 gigahertz channels to support a terabit data rates.  I am certain that they are not clocking those lines at a terahertz.

So for achieving those frequencies today, without being able to make an oscillator which can actually resonate at the specific frequency we want, say 610 terahertz,  we use matter to do a conversion for us.  Through the immense research and development done by the semiconductor industry we know that gallium arsenide emits light when its arranged in a particular configuration.  You start with a piece which is doped to be N type and a piece that is doped to be P type.  Doping is the process of adding impurities to the semiconductor material when its crystals are grown.  Dopants are usually boron, arsenic, phosphorus, and sometimes gallium.  The N type piece and the P type piece are bonded together and metal conductors are bonded on the ends.  When we pass direct current through the PN junction of the gallium arsenide material it emits green light.  So this chunk of rather specialized matter resonates at ultra high frequencies when stimulated with electron flow and emits green light.  This is your 600 terahertz oscillator.  By the nature of the gallium arsenide atoms simple electron flow causes it to resonate and emit light, like a radio tank circuit emits radio frequencies. 

The spectrum has different bands of energy, like radio bands.  It is not feasible to take a radio oscillator and make it resonate at 600 terahertz and expect light to come out of it.  The radio oscillator works good within its band, but it has limitations.  If you want to experiment in the light band of the spectrum you need a different physical apparatus.  The same is true for electrical current and gravity, both are in different bands of the spectrum and require a different physical apparatus to solicit the desired results.

Photons and Electrons are both energy, but they are not the same form and frequency of energy.  Lower frequency energy is relatively large quantum particles.  Higher frequency energy is relatively small quantum particles.  This is inversely proportional through the entire spectrum.  At the low end most energy is electronic (big quanta, like electrons) in nature.  As you move higher through the spectrum, past radio and microwaves there is a grey area where the electron and photon energies seems to be the same.  Then at the upper end of the spectrum the photonic (small quanta, photons) energies dominate because of their small size and higher velocities.  The wavelength of an energy is determined by its velocity.  The velocity of a quantum particle traveling through open space forms a wake in the ethers which we identify as its wavelength.  The frequency of an energy determines its velocity and a function of the velocity is the wavelength of the wake left in the ethers.

Well, I hope I didn't break your mind so early in the morning.  I'm all hopped up on coffee, and I like thinking about the quantum realm when I'm hopped up on coffee...

OK, Mo Later...

[z.monkey]

Howdy Art,

There are four pairs in an ethernet cable.  So at 100 megabits, you be getting 25 megahertz signaling in one pair times 4 pairs is 100 megabit signaling.  Tricky, they are, yes...

OK, Mo Later...

[sm0ky2]

ammount of data being transfered per cycle, vs number of cycles per second,

these values are not superimposable like that...

thats why they are represented as two seperate concepts.

[z.monkey]

Howdy Art,

Well, I know I am not making this up.  I just worked on a project where I had to troubleshoot a malfunctioning ethernet design.  I guess I'll have to go back and look at that more carefully.  Maybe I am thinking of the way the early 10 megabit ethernet works where slow clock rates necessitated transmitting 4 pairs at 2.5 megahertz.  I'll get back to you on that.

As far as the original topic, I think RedRiderNo22 shot his eye out...

OK, Mo Later...

[sm0ky2]

pc to pc, it uses all 8

but for raw data transfer:  i.e. - through a hub/switch, ect  ther are really only 2 pairs being used.
you can observe this through a crossover cable, ALL of the data is passed through 2 pairs of wires.

so i would assume that it is using parallel data-transfer through the 8 wires.