Being the obsessive compulsive person I am, I was forced to go look it up. Silicon is classified as a non-metal. It does have a low electrical conductivity, but it can conduct electricity. Here are some pages of interest I found:

http://hyperphysics.phy-astr.gsu.edu...lids/sili.html
http://hyperphysics.phy-astr.gsu.edu...intrin.html#c1

The process of 'doping' silicon (or other semiconductors) can greatly influence their electrical characteristics, but pure semiconductors are still capable of conducting electricity. That doesn't say much though, because if there's a large enough voltage differential across an object, I think it'd be tough to find any material that didn't permit electrical current.

The example of the oven is based on a different property of matter: thermal conductance. I'd have to crack open my old chemistry book to make sure, but terms like "specific heat", "heat of fusion", "enthalpy", and others keep dancing around in my head. It's been too long since I dealt with thermal properties to remember exactly which property or properties would directly influence the result of the oven experiment.

that example was just to show what silicon can do (and also explains why it would be used in computers, after all, a metal surface might melt/conduct to much electricty)

Nice reading Dark_Helmet. I remember sitting in class while i was in the Navy Learning about "doping" materials. It started off with forward biasing simple diodes and then went to transistor theory...ect..ect and ended up with digital ckts. Unfortunately, they made us spend 10 hrs a day 5 days a week for 6 months teaching us all this great stuff, but we never really got to use it. As it turned out once we got out in the "fleet" (what ever shipped we got lucky or unlucky enough to get stationed on)we would trouble-shoot down to either a board was getting an input, but wasn't putting out an output, or what ever, and we would replace the board, no component troubleshooting for us. And, good or bad, it even got to the point of not having to use a meter, just look at all the little lights..."yep those are on...that ones not...thats the problem."

I'm glad you found it interesting. It's a topic I like to consider I'm fairly knowledgeable about. Otherwise, my degree wouldn't be worth a whole heck of a lot. Not to mention my last job was part of a team designing storage controllers. You'd hear acronyms such as GTS and RTM thrown around a lot ("Go To Silicon" and "Release to Manufacturing"). There were others, but that's where I became quite familiar with manufacturing techniques of custom chips, and the process became something "real" rather than a few paragraphs in a crusty textbook.

Hehehe... the visual inspection method is a tried and true one. If there's a problem with a board, there's very little you can do about it. If you have access to the board's schematic you have a few more options, but not a whole lot more. You might be able to cut a trace on the board, solder a wire between two (or more) endpoints, or replace a socketed chip, but a board design is heavily customized and any modifications are likely to lead to errant behavior for some configuration(s). I had to do mods for some of our testing boards, and it consisted of cutting one trace and soldering two circuit points together. The board was useless for anything other than my specific test. And if it's something within the chip itself, well, you can entertain yourself by "letting the magic smoke out". For anyone not familiar, you let out the magic smoke by shoving too much electrical current through a device. The device emits a small puff of smoke, a very distinctive odor, and then refuses to do anything useful I did that by running 1 amp of current into the base terminal of a transistor in a lab one time. As reference, most transistors expect milliamps or microamps on that base terminal to operate correctly. I flipped the circuit on and *poof*.

Anyway, before I wast even more time reminiscing about the good ol' days of lab experiments and mishaps (never catch a capacitor tossed to you by the way), the point of all that was, the visual method is pretty much all you can do.

IIRC, AMD's official line on their 'Performance Ratings', is that they are used to compare the XP series of processors to the original "Thunderbird" cored chips.

The fact that they happen to closely match Intels GHz ratings is just coincidence.

And last I heard Intel were looking at using a performance rating system for their next batch of processors.

Silicon is a material that is somewhere between metals and non-metals with its properties. While all metals are conductors and (most) non-metals insulators, silicon (and germanium, and GaAs, and some other materials like e.g. diamond) is called a semiconductor. It has the interesting property that its conducting properties can be influenced by small additions of other materials and by applying electrical voltages. This makes it possible to make electrical switches that are controlled by electrical signals: the transistors in your electronics. I don't know much about quartz or CPU clocks, maybe google can help you.

To get back to the clock speed & big numbers issue: I have an interesting calculation for you. You know that nothing goes faster than the speed of light, 300 000 000 m/s. Now let's assume that electrical signals are very fast, approximately equal to the speed of light. If your clock is ticking with 1 GHz = 1 000 000 000 ticks per second, and everything in your computer has to receive that pulse before the next pulse can be sent, then the pulse has 1 / 1 000 000 000 second to travel. Calculating the distance: x = v / t = 0,3 m = only 30 cm. Real electrical signals travel slower than the speed of light, which will make things even worse. I guess that this is a big problem for making faster RAM (as long as it's not integrated in your CPU ).

cjp: there are speeds faster then light that (or at least what become) matter once traveled at

Great SciYro, way to blow my little monkey-mind! How did the energy turn to mass if, for it, time was moving backward? Seems I read about some particle that does move faster than light (Tachyons?) but such particles can never be used for communication.. you'd get your answer before you asked the question.

I vote for the standardization of the bogomip...

Nothing says 1337 like talking about how fast your computer can do absolutely nothing (please excuse me if I'm wrong about the definition)

GOBY: no way, theres no such thing as time, as far as particles are concerned, they know where or how fast they move, they know nothing of time, only we do as its a perception, based on space, and has thus has the same rules as space does, you cant move into the past ..... if you could, then we wouldn't even exist (or we could, but it would be one messed up world, and probably would make it so we (humans) never evolved into humans)

there is a way to communicate in a much faster time tho, just bend the dimensions so 1 particle will appear to be in 2 places in our dimensions, but in reality only exists in 1 place that happens to be in both places at the same instant (the overlapping of the dimensions)