Moore's Law.

[Rit]

Bigger? The'rye heading in the smaller direction, not bigger. Lol, lets make a CPU the size of a car.

Dang, wonder how hot that would get and what kind of cooling system you'd need.

 

[WeatherMan]

Sod that, I'd rather not have a substation outside my house powering my computer

A small office would need a power station
A large organization would need a nuclear reactor

 

[Thanatos]

Bigger? The'rye heading in the smaller direction, not bigger. Lol, lets make a CPU the size of a car.

Well, it's not like they couldn't add another square inch to the size of CPUs. That'd add another couple hundred thousand transistors to the chip, and we'd never even have to change the form factor of our Mobos. Cases stay the same size, CPUs are way faster, and everyone's happy!

 

[AntimatterAsh]

A lot of experts have predicted that Moore's Law will reach an end soon because of the physical limitations of silicon microprocessors.

Also, as you know, CPU manufactures are always striving to pack as much as they can on a die, which puts us down to 22nm (Ivy Bridge). Atom's are 0.5nm, so wouldn't we hit a wall at 0.5nm?

The reason im bring this up is because we aren't far off.

Quantum computing technology is still in its early stages, and I don't think it will ready in time when we hit that wall.

Just want to hear ya'll thoughts on the matter.

I don't know much about this but for quantum computing, instead of getting an on and off or 0 and 1, you can get an up, down and sideways, so 0 1 and 2. But to perform one 0, 1 or 2, you'd need over 11tb of RAM just to process the operation, meaning that you'd need hundreds of terabytes of data for a low end processor, but to be honest, I think the near future is multiple processors like servers. If it is found to make other parts smaller, why not add a second socket like in skulltrail boards?

 

[Ankur]

In recent technologies, the hardware part in computers are developing faster, software development is a bit slow, so what why will anybody need better processor for years?
You take a game as an example, it is the architecture, no. of CPU cores, frequency that matters most. The die shrink size only matters for efficiency, like leakage current is reduced in Ivy bridge due to the small 22nm 3D transistors.
Even if the CPU transistors goes down below 10nm, as seen on Intel's CPU road-map, there won't be much difference in processing speed, just efficient CPU. Even Super Computers like titan use multiple processors, so it won't matter what size, but just "how many" is the thing that will matter.

 

[AntimatterAsh]

In recent technologies, the hardware part in computers are developing faster, software development is a bit slow, so what why will anybody need better processor for years?
You take a game as an example, it is the architecture, no. of CPU cores, frequency that matters most. The die shrink size only matters for efficiency, like leakage current is reduced in Ivy bridge due to the small 22nm 3D transistors.
Even if the CPU transistors goes down below 10nm, as seen on Intel's CPU road-map, there won't be much difference in processing speed, just efficient CPU. Even Super Computers like titan use multiple processors, so it won't matter what size, but just "how many" is the thing that will matter.

Agreed. I do think that this introduction and craze with tablets is unhealthy for the processing industry though. All these processors for tablets are having to go back to the power and speed of generations back just to fit in the package.

In other words we are going back, not progressing fowards.

 

[Ankur]

Agreed. I do think that this introduction and craze with tablets is unhealthy for the processing industry though. All these processors for tablets are having to go back to the power and speed of generations back just to fit in the package.

In other words we are going back, not progressing fowards.

Actually for companies business is important and they want to generate more revenue every year, that is why they try creating new markets, like tablet, mobile devices market is quite good presently and also in the future, profit from this is quite high, imagine Tegra 3 without tablets, Nvidia like companies also going good with this.
So business point of view it is progressing forward . Technologically comparing it to the old 8086 and the Pentiums, yeah quite good.