My grandfather (incredibly smart man) brought up this question one day when we were talking about building circuit boards. We started out with vacuum tubes, then, in a stroke of brilliance the silicon computer chip was born. Computers have gone nowhere but up since then. But eventually, when we have quad cores, and 4 GHz GPU's, the expansion room will end. Silicon cannot carry us forever. Transistors, inevitably, will die. So my question to you, is what you think the next real innovation in computing will be. Not another core, not more transistors, not a 4 GHz CPU, but a real breakthrough. Obviously you will not know what exactly it will be, or you would be a very, very rich person. Just let your mind wander. Personally i think it will be some kind of manipulation of light. Light moves faster than electricity. And also, electricity has 2 possible readouts, positive and negative. Light has infinite possibilities. The wavelengths that can be read by a CPU are nearly endless. Sure when/if we started out it would only be 2 or 3 wavelengths being read, but the expansion room is enormous. This may sound impossible or exotic, and it is. Right now it's not possible, but in the future, think. In the 1950's people could not forsee a computer with less than a thousand vacuum tubes and weighing less than 1.5 tons. Imagine what can happen in the future. I want to hear good suggestions, who knows, one of us might be the next innovators. Just remember when you are on TV talking about how you did it, say, "A guy named max12590 on Hardware Forums started this thread about....."
Actually Max, people have been addressing this issue since the advent of the silicon wafer. As for your statement about light traveling faster than energy, that's actually incorrect. Energy moves at the speed of light. Whether we'll use light-based trace paths I don't know, although I couldn't see the benefit over the 1-atom-wide trace paths schientists are trying for right now. They're basically intending on shrinking the transistors and traces so small that PCs (video cards, etc) will be astronomically faster than previous generations. Also, scientists have figured out a way to send more signals besides just "open" or "closed" across those conductors, meaning another huge increase in performance and a fundamental change in how computing even takes place. We'd have to basically start from scratch if we were to implement a setup like that. -AT
I realize that this is not the first time this question has been brought up but I haven't seen it here yet. The thing is, that at some point, when a transistor is just a few atoms, it caonnot get any smaller. There has to be something beyond transistors. It is just waiting to be discovered, some currently unforseeable goal.
the new crossbar latch developed by IBM I believe holds the potential to make computers millions times more powerful than today's
One area with huge potential is quantum computing. Basically quantum computing operates with qubits, more than 2 different states of value. www.quibit.com
you know we should ahve a technical or Scientific forum for these type sof threads. They're interesting and could provide good discussions or flame wars etc .
Thats the thing, all wavelength can be utilized. As I said, at this point it is fairly unforseeable, but then again a 3 GHz computer was too at one time. Anyways, I just thought this would be a really good topic for people to discuss and maybe come up with some good stuff. It gets your mind working harder than replying to "My HD stopped working" and "I got a BSOD today".
I thought that was magnetic RAM that just held information in RAM so boot was faster, like flash memory. And the Pentium 1 in my room uses subatomic particles
it has memory pools and works incredibly fast, so far the largest spans are 16mb i think, its still in research, should be faster than XDR2 even, but once its done it'll be like 2008
Once we learn how to tap into zero point, vacuum, cosmic and ambient energy fields there's really going to be a beastly computer.
Hehe, somehow I'm not thinking that quibit.com holds a lot of promise when it comes to the future computing, either that or maybe I'm missing something. Anyone else checked out that site?
Sorry about the double post. Wow, I hope my next computer doesn't run Quibit OS 3.72. Those games couldn't touch Solitare.
I think the Cell processor idea show potential between the move to something beyond transistors, so like a cell processor version of the cross bar latch technology would kick much butt.
Yea, that would be nice. There are so many different possibilities, it may end up that many of them are used, as opposed to just one like with transistors.