Where do integrated circuits go once we hit 1nm? Grapheme is already off the table

Where do integrated circuits go once we hit 1nm? Grapheme is already off the table.

Other urls found in this thread:

techradar.com/news/computing-components/processors/this-powerful-1-000-core-processor-is-surprisingly-energy-efficient-1323679
twitter.com/SFWRedditVideos

optical circuits

Picometers

Magnets

Quantum Computers

700,000 picometers

The only material that might be stable at 1nm is carbon. Not even Germanium is stable at 2 to 3nm.

>implying we'll hit 1nm

quark transistors

Is this even possible? Can you manipulate quarks like you can with atoms?

I always thought quarks were a thing that was known, but not "editable".

lightbased maybe dunno

dont listen to anything retards on Sup Forums say about science

Carbon Nanomemes

>1 nm

not even close, 7 nm max

Vertical circuits, like cubes made of wires or some shit.

shut up, carbon is stable until 700,000 mememeters

If you can figure out what to do after 1 nm you'd make a shit ton of money.

Hell, if you could figure out how to get it to 1 nm you'd make a shit ton of money.

Physicist here. No. He's a complete idiot.

>Grapheme
more like graphmeme, am I right?

ayylmao circuits

People will realize that making thing thinner makes them more fragile and will start to make them bigger before we hit 1mm.

STACK LIKE 10 CPUS N TOP OF EACHOTHER! PROBLEM SOLVED!

ummm....no?

Moar cores, larger dies, die stacking, moar expensive CPUs, more robust cooling. It's the only solution.

Dark times ahead for computers.

...

Bendgate, for example.

>Grapheme is already off the table
And why is that?

once we can't scale down any more or squeeze out higher frequencies, ideally we'd work on getting more efficient instruction sets followed by better more efficient architectures but the reality is the following will likely happen:

>Moar cores, larger dies, die stacking, moar expensive CPUs

another thing that would be nice is ternary or higher computing, it's not like it hasn't been tried before

In the future, we will compute entirely in Memes. I suspect by the year that by the year 2100 we will have a processor capable of 1000 Teramemes/s.

>talking about 2100
>holy fuck bro even your kids won't live that much. calm the fuck down.

...

>Teramemes/s.

>not using the SI standard abbreviation Cp (Christopher Pooles) to measure mememagic

>Where do integrated circuits go once we hit 1nm?

Programmers will be forced to write non-bloated code. Immediate 5x speed increase in any app, but only if you upgrade to Win12, which people will be unwilling to do for a decade.

How many memes per second equals one Cp?

>Using cp
>Not using pp (poo pytes) created by bane for baneposting purposes and made public by harold finch which 70 billion people contribute everyday

They don't.

We've exhausted size limit.
We've exhausted clock speed.
We've multicored CPUs until diminishing returns.
We're still working on massive parallelziation (GPUs)

Computers as they stand are at their limits. It was a good run, Moore, but it's over. Intel pink slipping +12kEngineers is the death knell from people in the know.

What we will see in the coming years are new types of computer architecures, like multicored fabrics, that have more restrictive memory hierarchies, but can offer unique solutions to many problems untenable with modern CPU design.

Btw, quantum ain't the next thing. It will never be more than a graphics card like add-on to your computer. It will be just a FPU at most. No way you're running Quinix on that shit. If you think that, you don't understand what you are talking about.

Oh, forgot to add shit's been completely pipelined too.

Christopher Pooles measure the Elelctro Memetic waves ememeted mgically by Memes to influence bodies of mass. 130.45 Cp are roughly equal to 9001 N / 69 Hz.

Pretty sure we haven't exhausted everything. The photo-lithographic technology is nearing the end of its usefulness for silicon but there are other semiconductor materials. It's also theoretically possible to increase the size of silicon wafers. Increasing the size of the wafer allows for bigger dies and increasing the die size by even a few millimeters adds quite a bit. Ultimately we'll find a way to overcome these hurdles.

Switch to 0nm and prepare for -1nm.

Moving to more and more parallel workloads
techradar.com/news/computing-components/processors/this-powerful-1-000-core-processor-is-surprisingly-energy-efficient-1323679

AMD set their sights too far into the future with Bulldozer.

Quantum computing.

What i think will happen is that quantum cpus will eventually transform and will become billions if not trillion times faster than regular cpus. People will buy on demand pre-proccessed information from a massive database that has every possible answer. Ultra high speed internet will allow transfering the huge amount of data in real time. The era of the pc is dying out.

I see you know nothing about quantum mechanics or quantum computing.

do away with 8086 architecture for something better and have dedicated processing units

>nm

Totally man.
You should patent that right away.

Why can't we just switch to picometers? We went from micrometers to nanometers without issue.

(You)

Picometer processors are a trap? What? Will they cause a fuckin black hole to appear and break the fabric of reality or something?

Because when you get this small the electrons start phasing through the material. 7nm is believed to be the limit for useful ICs.

the kilocore is worse per watt than even Jaguar cores.

Optoelectronics, quantum junctions, and 3D stacked logic. If power per Xtor is low enough then self heating isn't a barrier to vertical density.

We'll start using magic and ditch technology altogether.

eventually the entire machine will be soc style and desktops will just be larger laptops. laptops will be like large tablets, large tablets will be like big smartphones and big smartphones will be like small smartphones. small smartphones will be like dumbphones and everyone will laugh at people who use them except no one will laugh at the drug dealers who can't get pagers anymore. as motherboards shrink and software emulation becomes more common we'll end up with credit card sized smartphones and desktops the size of a usb hard drive. then comes moot os 1.0 where everything is an anime.

But all of this already exists.

>mfw that's the joke

Nano meters aren't the smallest measurement, jackass. 1nm would be about 2 silicone atoms long.

Topkek
>Ememeted

>silicone

>silicone atoms

Jaguar cores are really efficient senpai

>2 silicone atoms long.

Stack them one on top of each other

3d manufacturing

Insulent fool, we Germans are stable at any size!
Unless you introduce outsiders, especially africanium from americium fucks our shit up.

where's my moot os? GAME SET MATCH, user

Who?

yes why
, also what about diamonds
before graphene diamond was being bantered about to replace silicon

The founder of 9gag.

Planck scale transistors and gravitational processing.

>not higgs boson based

The dominant force at Planck scale is gravity ;)

I'll be collecting my (You's) now.

...

...

tfw it doesn't change the title bar content

>Harold Finch

we just need to start building desk-sized computers again like the good old days