Have u ever wondered at a CPU's "wavelength"?

Wow guys. I was sitting in Physics class yesterday. We were studying waves and energy transfer. We had a test today (I aced it ) Assume that u can figure out a wave's velocity using the equation v =f ë , where v= velocity of the wave, f= the wave's frequency in Hertz (Hz), and the greek letter lamda ë is the wave's wavelegth. Assuming the electrical impulses ( electricity) travel at the speed of light (2.99792458 x 10^8 m/s) and that an individual is using just a 1.0 Ghz CPU, u can figure out that CPU's cycle's wavelength. Just rearrange the equation so that it reads ë=v/f . Plugging it like such:

ë = 1,000,000,000 Hz / (2.997972458 x 10^8)

We get an answer of 3.336 meters (rounded). lol. Its that just cool? A little chip still can produce an oscilatting wave that still just about 3.3 meters? Now just imagine a 3.06Ghz chip and u get a wavelength of 10.207 meters. lol. Now what about their periods? Well that another story:

T=1/f, where T is the period and f is the frequency. asumming the person is using a 1.0 GHz chip, the period (or time it takes for one oscil
lation) is 1 x 10^-9 seconds!!!. For you number freaks out there, thats .000000009 seconds. Now u throw in a 3.06GHz chip, and u get a amazing 3.268 x 10^-10 seconds (.0000003268 seconds). WOW. Modern technology is great, ain't it? The CPU is producing these fairly large waves but at such incredibly small time intervals!!! Mind boggling, ain't it??

P.S. If anyone finds an error, please let me know! Thank you!

Sorry dude. CPUs don't make waves. The Hz in the clock speed is cycles per second. The Hz in physics involves the time it takes a sinusoidal wave to leave one position and come back to the same position (cycles per second).

There is nothing in the computer that can be considered to have anything to do with 3 meters. The Hz is simply the time it takes for a signal to travel completely through the most latent path in a circuit.

This no longer even means that 1ns (for a 1Ghz computer) is the amount of time it takes for one assembly instruction to be executed either. It is the time it takes for one single stage in an assembly instruction to be executed. Most processors are now pipelined so they execute several instructions at a time with an offset of one clock cycle.

Actually, it's more like a stop-and-go situation: moving back and forth is for AC.

The clock signal is generated by a clock chip, in the form of a square pulse, varying between 0V and ... huh... 3.3v?

Also, the electron speed can be lower than the theoretical max, by as much as a full factor of ten: the mobo's traces (or CPU pins) are not superconductors.

You passed physics, congratulations! Now go on to electricity! (College physics 201 / 301).

The Hz in the clock speed is cycles per second. The Hz in physics involves the time it takes a sinusoidal wave to leave one position and come back to the same position (cycles per second).

Click to expand...

I think u canceled out your point.... U defined the same thing . lol. But i do beleive u are right about CPU's not producing actual "waves". Why then is is such a concern for CPU fracturing under extremely cool conditions (i.e. phase change cooling) duue to CPU oscillation? But i think I'm right about the period the cycle? Isn't a nanosecond a billionth of a second? So that'd be like 1 x 10^-9 seconds ? lo.. gotta love physics and computers!

SuperDave1685 said:

I think u canceled out your point.... U defined the same thing . lol. But i do beleive u are right about CPU's not producing actual "waves". Why then is is such a concern for CPU fracturing under extremely cool conditions (i.e. phase change cooling) duue to CPU oscillation? But i think I'm right about the period the cycle? Isn't a nanosecond a billionth of a second? So that'd be like 1 x 10^-9 seconds ? lo.. gotta love physics and computers!

Click to expand...

Read the post just above yours here. BigBen2k mentioned superconductors in passing. The only superconductors we have need to be super cooled (close to 0 degrees kelvin often times).

The colder you make your cpu, the lower its resistivity and the faster the electrons can move.

And as for defining the same thing, it is the same thing. Hz is cycles per second. It's just that a wave is a sinusoidal pulse. Clock pulses are different.

If you put the clock pulses down a long wire, each pulse (well, the peak of the pulse, at least when it is first reached) would be that far apart, right? What would be the difference between a clock cycle and a really funcky wave function when comparing graphs? Sure, the clock cycles would be rather squared off, but they would follow a period in the same manner. And Hz are Hz no matter where you look, all it is is 1/s.

i think you made a mistake in your calculation somewhere. a 3GHz pulse should have a shorter wavelength than a 1GHz pulse. also if i remember from my digital circuits classes the size of a cpu has to be an exact multiple of the wavelength or something like that. and the frequency has to be droped way down anytime you go outside the cpu or all kinds of bad things happen. anyway it is cool to think about.

What happens is more like a transient response in a transmission line. One would use Fourier wave analysis to find the max length of the buss lines and the optimum impedance to minimize ringing.

Check your math, you should be getting about 30cm for 1Ghz and 10cm for 3Ghz.

What worries me is you've got this porcupine of a half wave radiator (5cm high heatsink) on a 3G CPU and a possible tuned cavity (accidentially the right shaped case) with a possible unscreened exit window (unshielded 3.5inch drive bay with just a plastic cover over it) pointed straight at your head (when your case is up on the desk) I think there are certain situations where you could come close to microwaving your brain without realising it.

Tip; if your Cold cathode flourescents glow when the power to them is off, worry!

Road Warrior

Electrons do not travel at the speed of light in silicon or metals, so you need to use a different velocity to get your wavelength. Look into electron mobility in silicon.

I'm seconding Doc's views. Also, 3x10^8 m/s is the speed of light through AIR, which may or may not be the medium in which it's traveling through.

Edit: BTW, great forum, really gets the brain juices flowing!

RoadWarrior said:

What worries me is you've got this porcupine of a half wave radiator (5cm high heatsink) on a 3G CPU and a possible tuned cavity (accidentially the right shaped case) with a possible unscreened exit window (unshielded 3.5inch drive bay with just a plastic cover over it) pointed straight at your head (when your case is up on the desk) I think there are certain situations where you could come close to microwaving your brain without realising it.

Click to expand...

Man... That really sucks to think about that!! But how much energy is actually released? I mean, I don't particulaly like the idea of sticking my head in front of a 1KW microwave ham radio antenna, but a CPU is FAR from 1KW, and probably still way less then the exposure limits set by the FCC.


I can hear my P-100 on the radio BTW
JigPu

JigPu you dont have to worry to much about being radiated by microwaves from your PC. If you have a metal case this will ground them and your brain wont get fried. Besides that, a CPU is a crappy antenna and most of the power going into it is coming out as heat, not as microwaves.