Moore's Law is dead

[Anonymous No.10756161]

What will take its place?

[Anonymous No.10756198]

Technology has stagnated the past 10 years. Difference between 2000 and 2010 was huge. Difference between 2010 and 2020 are insignificant.

[Anonymous No.10756769]

>>10756198
Because silicon has reached its limit and we need a replacement. Next decade marks the rise of quantum computing.

[Anonymous No.10756770]

>>10756769
>Next decade marks the rise of quantum computing.
Next decade marks the rise of computers with massively increased error rates.

[Anonymous No.10756772]

>>10756769
>Next decade marks the rise of quantum computing.
https://www.youtube.com/watch?v=mIULkvTGk2U

[Anonymous No.10756784]

>>10756161
>The 2013 report wasn't even right about 2014
Lmao what the fuck is going on at Arm?

[Anonymous No.10756795]

>>10756772
>uploaded feb 2018
oof this didnt age well

[Anonymous No.10756810]

>>10756161
infinite growth in a finite universe It's a law ?

[Anonymous No.10756814]

>>10756161
https://youtu.be/DZCm2QQZVYk

[Anonymous No.10756824]

moores law was always a meme since there's obviously there's more to chip performance than simply number of transistors in an area. right now a trend in academia is looking at 3d integrated circuits (stacking layers of transistors on top of one another), which will continue moore's law. the performance benefits are not because there's more transistors, but rather because system architecture can be streamlined in a true 3d chip.

but it is true that traditional silicon scaling is really hitting the wall. there will likely be a future in looking at new materials with higher mobility to replace silicon channels.

quantum computing is never going to replace conventional digital computing except in a few specific problems. why the fuck would you need a quantum computer to calculate 1+1? even if room temp magic perfect quantum computing existed, it would just be a small specialized block on a CPU die.

[Anonymous No.10756859]

So now that Moore's Law is dead does that mean the Singularity will never happen?

[Anonymous No.10756882]

>>10756161
Reversible computing
https://en.wikipedia.org/wiki/Reversible_computing

[Anonymous No.10756884]

>>10756161
The fact that there isn't a consensus is a very bad sign.

[Anonymous No.10756887]

>>10756859
It's just a mathematical abstraction.

[Anonymous No.10756898]

>>10756884
It just means we're at an interesting point in computing. It's actually a pretty cool time because we can't just sit around and wait for more performance to come in the form of smaller transistors that we can clock faster. We actually have to innovate in areas like computer architecture and more specifically memory systems. Before, if you wanted to do something fancy with DRAM, it meant using last year's memory in the new product -- that sucked because the memory technology has gotten faster by the next year so whatever performance gains your new cool thing would have given is outweighed by the fact that its probably going to be too expensive to incorporate into other products.
Now though, memory technology changes less from year to year - meaning there is more time to innovate. The Hybrid Memory Cube was almost a thing and it was really awesome (RIP).

[Anonymous No.10756909]

>>10756859
We haven't even scratched the surface of what's capable with software. Even if processor speed completely plateaus (unlikely), innovation in software will continue to boost computer output exponentially.

[Anonymous No.10756918]

>>10756161
In quantum computing I have seen it's called Nevan's law or Rose’s Law.

[Anonymous No.10756962]

>>10756859

Moore's law is not the first paradigm in exponential computing power.

Also, Moore's law only refers to the doubling of transistor count not the doubling of exponential computing. There is a subtle and important difference. Specialization and other techniques in certain components of the computational components in computers have picked up the slack. For example multi cores and 3D stacking have mitigated effects of moore's law tapering off. This comes with a disadvantage of more research efforts and complexity. We only need this to last until computational systems can match the human brain. Different computational paradigms will emerge, only the efficiency and growth rate is what is in question.

[Anonymous No.10757015]

>>10756161
Balanced Trinary – {-1,0,1}
Trinary Coded Binary – {T,F,T}
Combining binary, ternary and quantum computing by way of the qutrit could be a fairly useful philosophy in designing software for quantum systems.
I read an article a few days ago about a group at the University of Chicago showing promise with their application of the qutrit in algorithms for quantum computers.
It's such a very exciting time to be alive.

[Anonymous No.10757047]

We can't make feature sizes smaller, but we can use what we have better, like going from planar to 3D.

We can also walk away from the idea of a CPU and go for full reconfigurable computing. Coarse Grained Reconfigurable Arrays are going to be the big thing in 2020 - 2030.

[Anonymous No.10757819]

>>10756161
Only because of greed. You see, it creates incentive not to wait years when you buy now, allowing you to jack up the price and force consumers to take up loans on stuff like cell phones.

[Anonymous No.10757865]

>>10756161
Your mom's law. Double in size and half in power every year.

[Anonymous No.10758869]

>>10756859
yes

[Anonymous No.10759854]

>>10756198
/thread

[Anonymous No.10759863]

>>10756161
Multicore and distributed computing.

[Anonymous No.10759870]

It's not dead for computers in general, but it has tapered off for single cores within CPUs. That doesn't mean you can't just add more cores. The trick is getting the cores to work together. Need lots of RAM speed to properly use multiple cores. DDR5 RAM will allow for some pretty decent multicore gains.

[Anonymous No.10759893]

>>10756161
There will be more competition in hardware design, security and optimizations.

[Anonymous No.10760058]

>>10756161
Optical computing

[Anonymous No.10760064]

All right here's an idea:
How about just fuck it and leave it as it is? It's enough for everything.

[Anonymous No.10760392]

>>10760064
It's not enough for singularity, anon.