Why processors are small?

smaller dies mean higher yields

/thread

My girlfriend used to say size doesn't matter. I guess it takes less space and that's all.

It also means that information takes more time to travel, therefore the processor must operate at a slower frequency to properly function

Holy kek user. Learn to use a search engine. It's in the fucking sticky.

Attached: 1526670658959.jpg (600x536, 71K)

but electrons travel at nearly the speed of light

but the speed of light is really really fast

The distance electricity travels in a nanosecond is 11.8 inches

to allow planned obscolescence by making shipping costs lower

>but electrons travel at nearly the speed of light
Yep, in a vacuum. Not in a circuit.

My math is likely wrong here, I'm not a math person but

At 4GHz, that's a clock cycle every .25 nanseconds. So .25*10^-9 seconds. That's a very small number of seconds. Light might be fast, but the clock speed is so high that in the time it takes a 4GHz clock cycle to complete, light can only move ~7.5 centimeters. And electrons are only *almost* that fast so things are actually even tighter that that.

i said nearly i know its not as fast i think its like 80% the speed of light still very fast

Yeah see that's even worse, 80% is a big difference. That's barely 6cm. 2 inches. Any bigger than 2 inches and you're breaking the laws of physics to get from one end to the other within a clock cycle @4GHz. That's why they're not any bigger or faster, that's the laws of the reality telling us to back off.

dude chill im just saying electrons go fast i always wondered that question and now its awnsered

Lol you got the answer like 5 times already and kept saying "but light speed FAST" every time

i said it twice some other guy said it the other times

Can we use light to send data in the future though? Would that mean a significant improve in performance, or only "20%" ?

s/5 times/twice/ and my post is still basically unchanged dude

>big=good

Attached: 1535580550273.png (707x786, 141K)

>laws of the reality telling us to back off.
quantum entangle the cores fuck the speed of light

20%'s pretty significant but I don't think anybody work on light-powered computers. I'm no physicist but light and electricity aren't really interchangeable.

>pins
God I'm so glad we've pretty much switched to contacts. Who thought pins were a good idea?

alright go be angury then dosent change anything you win is that want you want finally something if your meaningless life that went your way you have won your a winner finally you have become chad nice one

I'm

Not

Angry

>:(((((

the fact that you're reading his posts with an angry tone says more about you and your state of mind, tbhfamily. he's not being a dick at all.

>light and electricity aren't really interchangeable.
what are photovoltaics?

your a winner

calm down son. have a sip.

It's not about how big it is, it's about how you use it.

i dont need a sip you need a sip

that's not how that works

What like solar panels? Like I said, I'm no physicist but how are you going to implement logic gates and transistors with light? Convert to electricity and we're not at LIGHT SPEED anymore

>80% the speed of light still very fast
Speed of light is about 300 000 000 m/s.
80% of that is 240 000 000 m/s. That's a huge difference.
Keep in mind that your CPU works way faster than mere seconds.
Assuming your CPU is 3.5 Ghz, that's 3 500 000 000 clock signal per second.
Which gives you 0.0685 meter per clock signal, which is 6.85 cm (about 2.7 inches).
So no, you can't build a huge and fast CPU.

>but electrons travel at nearly the speed of light

Actually, electron drift is extremely slow. Electrical fields move at nearly the speed of light.

Jow Forums's poor understanding of electronics is showing.

its not actually true thou. the reason they are small is they where developed in the cold war where they thought they would have to be small to fit inside military shit.

you could make a huge one. just look at GPUs you can use like 7feet of pcie extension cables and still get same performance same thing with DDR extenders which actually exist. only thing distance effects is voltage not speed.

Because it takes more time to send data (electricity) over larger spaces. This is on a huge scale but I'll give an example. I want to send an instruction from point A to B. If my processor is bigger then points A and B will also be bigger. If we shrink the processor then points A and B will scale down but that gap becomes smaller, therefore the time it takes to close that distance becomes shorter.

Does that make any sense? So basically a bigger processor is slower. And no, it wouldn't be cooler if you still have the same amount of stuff spread out over a larger area. It would still pull just as much power.

Attached: 1537909404786.gif (250x400, 23K)

>just look at GPUs you can use like 7feet of pcie extension cables and still get same performance same thing with DDR extenders which actually exist
what the fuck are you talking about
why do you think latency isnt an issue?

AMD agrees with you, OP.

Attached: Amd-ryzen-threadripper.png (725x512, 90K)