As most of you know, we are about to hit the limit on how small silicon transistors can get, meaning we've almost reached its full potential.
With no clear alternative in sight, we may be stuck with today's graphics for quite a while. Or will there be a new type of architecture before we hit the limit? Are we going to see an even bigger stagnation in video game visuals in the next decade?
>inb4 quantum computers
>>320410802
Just make graphics cards physically larger. Back in the day, everyone had full towers.
>>320410802
>As most of you know, we are about to hit the limit on how small silicon transistors can get, meaning we've almost reached its full potential.
C I T A T I O N N E E D E DYou don't have one because it's not going to happen for like 20 years, fuck off
>>320410802
Do you have a single fact to back up that statement? Citation? From someone that isn't just an advertisement.
when there's money on the line innovation will come in no time
>>320410802
Quantum Computers in the current way we know are literally unable to play games.
They're basically like that space ship in Hitchhiker's Guide to the Galaxy that does everything, ecerywhere at the same time but just not the one thing you actually want to do.
It's true what op said.
But in 1961 people thought they knew the size limit of a transistor and now we have 1/100 of that size.
So I wouldn't be worried.
>>320411313
>>320411421
While we might not actually reach the limit for another decade or so, Moore's law is no longer in effect, as of 2015.
>>320411896
Back then they were limited because they didn't have a precise way to make such tiny transistors. Now we have the technology to make the transistors as tiny as we want, but if we make them too tiny they'll be useless because an electrical charge will travel through the transistor whether the gate is open or not.
Back then out own technology was the limiting factor, but now it's down to physics which is a much bigger obstacle.
>>320412227
>Moore's law is no longer in effect, as of 2015.
Says who
>>320412854
Intel. They said the number of years for a doubling in the number of transistors is up to 2.5 or more and it used to be less than two.
>>320411195
This. There's plenty of room for more cores and better SLI support.
>>320412227
>Moore's Law isn't in effect
>I don't know about the topic so I will write some stupid lies
No, what it's true is that the rate at which processors are doubling their maximum processing power is slower than years ago, and it is also true that clock speeds and W hasn't increased, but as long as we can still add more cores and further reduce the size of transistors, we're fine.
Only when transistors are 5nm, then we will be stuck with a fixed IPS.
>>320413703
Moore's law said it would take ~2 years to double the processing power. It now takes 2.5, which is significantly more. Therefore, it is no longer in effect.
God I hate explaining things to children
