Gigahert Proscessor ?? I know something better

[batmeat]

Ok, think about this. Processors will eventually reach a limit in speed, whether it be because of the .18 (or even smaller) micron technology, hardware, heat, etc. If you could design a processor that utilizes light, possibly in the form of a laser, data transfer would be nearly instantaneous. The problem is setting it up.

[jad1097]

It is being done. IBM is doing something like that. http://www.zdnet.com/pcweek/stories/news/0,4153,2453929,00.html


I misread the artical http://www.sysopt.com/forum/frown.gif THis is something different.

[This message has been edited by jad1097 (edited 03-23-2000).]

[jadison]

I read about this a few years ago...what they might have to result to is having molecular processors, where the processor itself uses parts that are only a few atoms thick. I read about this in Popular Mechanics.

I'll try to find the issue...until then, keep brainstorming!

-jd-

[batmeat]

IBM huh, Gee, I thought I was on top of things.

[Apostle 83]

Didn't i hear about a month ago that (place name here PLEASE!) had shattered the theoretical limit of chip pathway size by 5ox or somethin?

[jadison]

Here's some more food 4 thought, it came from Popular Mechanics Issue 6/99:
"A new low-cost technology could replace microchips with spherical crystals.
Ball Semiconductor(www.ballsemi.com) says it has fabricated a transistor on a 1mm silicon sphere. To accurately etch circuit elements in 3D, the Allen, Texsas, company developed a mirror with 45 facets to transfer a circuit pattern while compensating for spherical distortion on the silicon surface.
The chief source of savings is the ability to make pure silicon globules inside small air-tight tubes, instead of clean rooms.
Future chip plants could cost one-tenth as much as today's facilities. Plans are under way for full-scale production some time next year."

It's a bit out-dated, but they said "sometime next year", so we should be seeing something soon.

-jd-

[Smokey]

In the summer of last year, I heard that HP and some other Californian company had created a nano processor, which uses atom-sized parts and stuff. Apparently, it whups whatever we have now.

[Donkey]

There is quite a lot of scientific research at the moment into trying to use single electrons trapped into certain arrays to achieve logic bits i.e. one config is 1 other is 0. This can also be done with small magnetic domains as they will either orientate in one direction or another. Hitachi and others are certainly working on single electron SDRAM at the moment i'll see if i can dig up some of the papers and post them.
THings are certainly going to have to change soon if we still want to make thing smaller and smaller. And i think we are living in a very exciting time as some damb clever technology is going to be around sooner rather than later i reckon.

[chipbgt]

Wouldnt it be just as easy to start hooking up four (or 5, or 8, or 10) processors together and runnning them in a sort of SMP mode? like 4 800 mhz all running together. Or is this way oversimplified.

[RobRich]

New solutions will be needed for cpu development. IBM has pushed the theortical limits of production to less than .1 micron, but that's many years off. Right now, it looks as if cpu's will have to start perferming more per clock cycle than they currently do. On suolution is the next gen 64 bit cpu's (Merced and K8), but they will have to catch a larger market than chips like the 64 bit DEC Alpha. Another stategy is to implement more proccsing units per chip, i.e. a chip with several multiple integer and floating point units.

Everything upto here is based in regular computing technology. Other tech is rapidly approaching, though. Quantum computers are most likely not far off if their development continues at its current pace. They function miilion's of times for effectively than our current systems since their circuitry can exist in both an on and off state at the same time (confusing yet?). Also, genetic computers are already being designed. I'm not advocating these systems, as I find building a computer from DNA rather troubling. They are to be able to manipulate fata by chaning key DNA structures and building new nerve links, just as the brain does.

[narayan]

I have heard about the DNA computer and it is a fascinating science. It does raise some very interesting questions. I read in the local newspaper here a couple of years ago that a 1000 gallon tank of DNA would outperform all computers in the world!