Computer speed increases will eventually be impossible

[Humodour]

Physicists determine the smallest possible communication distance/time and extrapolate how long it will take to get there. If Moore's law continues at its current rate (and it has so far), we'll hit the physical wall in the next 75 years.

 

This isn't the same as the thermodynamic limit or such - those are physical barriers which smart design and engineering have so far been able to overcome. This is instead a sort of absolute, like the speed of light (in fact it's tied intimately to the speed of light).

 

Won't that be a novel concept? Computers that can't go faster...

 

http://www.physorg.com/news174750105.html

[Hiro Protagonist]

I wouldn't mind a computer that runs at the speed of light.

[Morgoth]

Maybe then Crysis will finally run smoothly maxed out?

[Rosbjerg]

Foolish if you ask me.. They are assuming that we will continue to use the current systems indefinitely.

 

One of the comments on the page you linked to gave a good suggestion, instead of using binary we could simply use the colour spectrum, which would allow a much wider varity than 1s and 0s and thus a faster computation.

[Walsingham]

I thought quantum computing was instantaneous?

 

Also, what about parrallel processing?

[Wrath of Dagon]

Yes, most of the speed increases now are from parallel processing. It's difficult to increase clock rates further because of heat and latency.

[213374U]

One of the comments on the page you linked to gave a good suggestion, instead of using binary we could simply use the colour spectrum, which would allow a much wider varity than 1s and 0s and thus a faster computation.

And if you read further down, you can find a rebuttal.

 

In the end, you are still sending data packets across a given distance. The time required for that packet to reach its destination cannot be infinitely small. Using colours means working with photons, which like the rest of EM radiation, travels at the speed of light. As Krez says, that is the true limit, and there's no way around it, that we know of.

 

 

I thought quantum computing was instantaneous?

 

Also, what about parrallel processing?

From a theoretical perspective, "instantaneous" makes sense only if time itself is assumed to be continuous. edit: and even so, I'm not too sure...

 

http://en.wikipedia.org/wiki/Quantum_time

 

Parallel processing makes computers faster the same way more bandwidth gives your connection a lower latency... that is, not at all.

Edited October 19, 2009 by 213374U

[Walsingham]

You mean one processor doing ten calculations is the same speed as ten processors each doing one calculation?

[213374U]

No, I mean the speed at which one processor does one calculation doesn't depend on how many processors you have sitting next to each other.

[Walsingham]

No, I mean the speed at which one processor does one calculation doesn't depend on how many processors you have sitting next to each other.

 

Oddly enough, neither did I.

[213374U]

I see. So, then the answer is "yes".

[Calax]

We may not be able to make the processor do calculations faster, but isn't the duel core design made so that it can do multipul calculations at the same time?

[alanschu]

Yes, so in that regard, you will be able to do more calculations.

 

But I think this discussion is more to do with the technology inside of a processor (i.e. single core).

Edited October 19, 2009 by alanschu

[Gorth]

Entangled Photons... a computer so fast that you get the result before you execute the program

[Lare Kikkeli]

Whats news is that we'll hit the limit in 75 years, not that there's a limit.

 

Actually we won't get very near light speed in computation since accelerating any object with mass, no matter how small, to light speed will take more energy than there is available in the visible universe.

 

Quantum computation will solve the issue with transistors, but it still won't let us break the laws of physics.

 

Just came up with an example: you have a 1 kg block of something, say hash, and you halve it. then you take one of the halves and halve it, take one of the halves & cut, then half etc. Every time you divide your block you need to use more energy. you can never reach 0, since your energy consumption goes up with every cut. at some point, as you're getting closer and closer to 0, your energy consumption gets higher and higher, and the max it can reach is infinite. since there's a finite amount of energy in the universe you can never reach 0. it's same with light speed, the closer you get the more energy you use. getting to light speed would take an infinite amount of energy so it's impossible. you can get close, but at some point the energy consumption gets so big that it's basically unattainable.

Edited October 19, 2009 by Lare Kikkeli

[Gorth]

Which is why completely new thinking is required...

 

As long as there are people who think we are limited by the speed of light, we will be limited by it.

[Hiro Protagonist]

We just need someone to build a Tachyon Computer.

[Lare Kikkeli]

Which is why completely new thinking is required...

 

As long as there are people who think we are limited by the speed of light, we will be limited by it.

 

That's against the theory of relativity, which while not perfect (it all breaks down on the quantum level) has been spot on so far. I'm not sure how much physics you know so I'll tell ya, going faster than light would be pretty damn weird. So weird that our 3 dimensional brains couldn't understand it.

[Walsingham]

Which is why completely new thinking is required...

 

As long as there are people who think we are limited by the speed of light, we will be limited by it.

 

I agree. It's nonsense. Gravity is already known to deliver effects at beyod the speed of light. Therefore light is not the measure of speed.

[Lare Kikkeli]

Which is why completely new thinking is required...

 

As long as there are people who think we are limited by the speed of light, we will be limited by it.

 

I agree. It's nonsense. Gravity is already known to deliver effects at beyod the speed of light. Therefore light is not the measure of speed.

 

You're right, gravital waves (which we haven't observed yet, it's all theoretic) would have to be faster or at least as fast as light. But would we be abled to send information with them? Can we even observe them? Can a quark exist by itself? There's so much stuff we don't know, but assuming everything is possible since there's so much we don't know is a logical fallacy. Can effects be delivered faster than light and can WE deliver information faster than light are two completely differerent question.

[Gorth]

Try googling for Entangled Photons and University of Vienna

 

They test faster than light communications. They have estalished that it works. They haven't found out why it works yet, but once that hurdle is overcome, there might be some new quantum leaps (pun unintended) on its way in science.

[Lare Kikkeli]

Try googling for Entangled Photons and University of Vienna

 

They test faster than light communications. They have estalished that it works. They haven't found out why it works yet, but once that hurdle is overcome, there might be some new quantum leaps (pun unintended) on its way in science.

 

I'm aware of that experiment, but it still will not allow faster than light travel.

 

Instantaneous communication by means of quantum entanglement is actually impossible because neither side can manipulate the state of the entangled particles, they can only measure it (see No-communication theorem). This fact means that if you measure one particle you cannot infer anything meaningful about the observers measuring the other particle, except you know what state they will measure, or have already measured. Thus causality is preserved.

 

Faster than light travel, as far as we know, would break causality.

[alanschu]

I have a question since it's been a while since I took physics, but what exactly are people referring to when they say they refer to "very near light speed in computation?"

 

 

 

 

(as an aside, I'm not sure where the speed of light came into the discussion. The article didn't say that we'd max out our speed because of the limit imposed on us by the speed of light, but rather that there was a limit much like how the speed of light is a physical limit).

[Lare Kikkeli]

I have a question since it's been a while since I took physics, but what exactly are people referring to when they say they refer to "very near light speed in computation?"

 

 

 

 

(as an aside, I'm not sure where the speed of light came into the discussion. The article didn't say that we'd max out our speed because of the limit imposed on us by the speed of light, but rather that there was a limit much like how the speed of light is a physical limit).

 

Because the speed of light is (according to the theory of relativity) the ultimate limit of transferring information. No information can be sent faster than light or it will break causality. The information, say the result of the calculation made by a processor, can not reach us faster than a photon because that would break the laws of physics. Neither can the calculation itself happen faster than the speed of light.

 

Now that I read that article again it does seem to put a limit on how fast a single processor can be, separate from light speed. I doubt it's even near the physical limit of transferring information.

Edited October 20, 2009 by Lare Kikkeli

[I want teh kotor 3]

Both saddening and great. I guess it makes sense, though. I'll assume it is a speed of light thing, and, in that case, it's a logical step.

 

We just need someone to build a Tachyon Computer.