Considering that the relative size of computers has shrunk from a two room tube-driven adding machine computer in the 1950's, to a common set up for the desk top which can exchange information with other computers around the world and store data in the quantities of 80 gigs and more for personal use a mere 50 years later, what will be the threshold for minimizing the size of the current P.C. A palm pal is not quite as strong as the PC in overall prowess, but who knows something about what the future of computer development holds in the next ten years? Any information would be very well appreciated.

King,
your question brought up memories of my electrical engineering class in college that i didnt like too much, but i remembered some of what they taught me,

so here it goes.

Moore's Law is probably the most important word in regards to that question. It is an equation that relates the processing power of the cpu over time.

Moore's Law tells us (basically) that if you take whatever information processing power that we have today, and multiply it by two, then that is what will be standard 18 months from now.

basically, the amount of transistors that we can fit on a certain size silicon chip will double about every year.(the amount of transistors tells us how fast our processors will be)

(up until a point a decade or two ahead at which point the transistors will become so small that we will have to find a new way to design them, i.e. quantum logic)

another governing factor will the the size of the hard drives, but they too as you mentioned are getting incredibly small.

It should get to the point in a couple of years where you will be able to do everything on your palm pilot that you can do on your desktop pc right now.


if you are interested, there is a really long website that talks about Moore's Law and the evolution of computer technology:
here

hope this helps,
-Chris

The computer components can be shrunk to near invisibility, but the human interface cannot. So long as human fingers and eyeballs need to be involved in the process there's little motivation to make the rest of the computer smaller than those components already are.

If speech recognition technology keeps advancing, we can eliminate the bulky input devices. Cochlear (in-ear) implants are already available for audio output and could serve as input as well. How about a display that stimulates the retina directly?! Surely such a device will someday exist.

If the input and output devices are part of your body, the rest of the computer is already awfully small already. Could easily fit the chips into the many voids found in a human body. Hollow out a little section of the bone in your arm, for example.

All that's left is the power supply. The power drain from a cochlear device is very small. Same would go for direct retinal stimulation. If input could be gathered by monitoring brain waves (already happening in the laboratory, and there's a video game controlled by brain wave feedback, can't recall the name), then very, very little power would be consumed so the power supply could be miniscule. Heck, something that small might be powered by the energies already present in your body.

I'm convinced that in the next 10 years humans will be augmented in this manner by computer systems integrated directly into their bodies. It's a natural outcome of 1) relying on computers for more and more and 2) shrinking components.

-- Craig

I think we're coming near a plateau in the areas of size and speed advances.

The advance that made the leap from electron tubes to transistors (the development of solid state components) has not yet been matched. Every advance we've seen since then has only been an improvement to solid state. There are only so many variations and most have been tried.

The real difference between a tube and a transistor (for example) is that a transistor is nothing but a piece of germaneum doped with arsenic to make it do the same thing. So you can potentially have a tiny chunk of "stuff" rather than a mechanical device. The little chunks can't be any smaller until we figure out a way to make the leads smaller. After all, they have to be mounted and connected to be useful.

There are about a half a billion different ways of making doped components, but they're all the same idea and have similar limitations, just different uses.

The next real leap was the invention of the integrated circuit, which is really nothing more than a bunch of tiny solid state devices inside a case. Further advances in processor chips aren't much more than putting in more transistors, arranging them differently, or making them smaller, and they can't get much smaller.

Future advances that might be handy:

A portable display unit that can link to your stationary pc somewhere in the world (via satellite or booster station) to access any of your files and another port to hook up a printer. Would be nice to plug in your cell phone, which by the way has the domestic or global calling plan. Convenience stores can put in networked downlink stations for your computing needs as easily as ATM's with modems.

Solid state devices that handle high power (tubes are still used in things like airport radar and televisions). As far as I know, only TRIACS and SCR's can do that now, and they're not much more than high-speed switches. This would be great for science and the government, I don't know what a private person would need it for.

Optical devices smaller than LED's and photodiodes might lead to something smaller than an integrated circuit--like a tiny plate with a light grid transferring commands via the air to the next component. Coupling circuits already use light-emitting diodes to transfer data to photodiodes, but they're still a little bulky and handle only a serial stream.

If we didn't have to use cables and wires as signal mediums, we'd make some impressive advances. Air is a very suitable medium, but it only works for virtually unguided signals like radio waves. Remote controls and very sophisticated missile guidance antennas (weighing several tons) are among the few pioneering devices. If you can imagine something like a finely crafted crystal with a hundred or so facets and an internal light source blinking commands in assorted directions to non-reflective components made of the same material as photodiodes.

Or what if we weren't restricted to binary? Components that understood more than just on/off would be nice.

Someone has to be working on these things somewhere.