There are things better than graphene. But pure electronic drift velocity isnt everything. Silicon is just really easy to work with, so traditional computing will likely continue to use it.

Is it not a question of use?

I'd say silicon for static and carbon for mobile electronics

>>7750808
No, they don't switch the way we need them to.

That said we can combine them with current systems to dramatically improve electrical and thermal conductivity, which will translate into significant performance increases. But those nice gains will not be the huge revolution people talk about.

>>7750938
>improve electrical and thermal conductivity, which will translate into significant performance increases

In what sense, power consumption and being able to run at a higher current?

>>7750944
Better electrical conductivity lowers power consumption and better thermal conductivity keeps chips cool allowing us to push for higher frequencies, which translates to faster processing. They all tie together so closely that I usually treat them the same, as lower power consumption also means less heat allowing for higher clock speeds. Obviously such things are not a positive feedback as there are limits, this just expands those limits given us more room to do things.

How these gains manifest is subject to how the designers exploit them. As they could keep processing performance the same and get huge drops in power consumption or push for decent gains in clock speed to improve speed or some mix. Given the markets I suspect they will go more for better power consumption as speed gains are not as lucrative as they use to be and many such speed gains can be made by fixing the poor software everyone uses.

Overall it expands the range of a what a system can do, so we will get better systems. What kind of gain is more in the hands of who builds it.

>>7750854
What do you mean by that? Workstations vs phones?

>>7750808
is that cosinus

>>7750808

No, we've spent decades over engineering Si. We're not going to do the same for C

>>7751074
We've spend decades using bronze swords, then made steel swords. Carbon (especially graphene) is being heavily researched!

>>7750854
>I'd say
never listen to people who say this.

>>7750808
Your Gif sais it all... Graphene has no band gap. --> it is no semiconductor --> worthless for electronics...

>>7752418
To be fair, it's not like carbon only exists in the form of graphene.

Diamond has a fantastic bandgap and high thermal conductivity. It'd be great for high-temperature electronics.

>>7752418
CNTs can be both semi conductors or metallic

>>7750808
I don't know much about the properties of carbon electronics, but I've been working with semiconductor industry for a while to comment about them.

In short, if the answer to these questions are yes, definitely they will replace them:

- Can they be manufactured reliable on a single substrate?

- Can they be integrated easily without throwing away the advances in lithography?

- Can they deliver higher performance? It could be fmax, current density or anything.

If the answers are yes, people would switch instantly with no regrets. People like to think that awesome devices make awesome circuits but it does not really work that way. Awesome integration capability is the driving force behind silicon.

An example would be ADC design. There are bipolar processes with devices going up to hundreds of GHz, but people still design their time interleaved ADCs in CMOS because they can throw a big CPU at all their reliability issues that way.

>>7752449
Basically, when you have to design and fabricate *billions* of nanometer-sc scale switching elements per chip, and then make millions of those chips reliably and cost-effectively, integration qualities trump *everything* else.

I will leave this here, as it helps address the elephant in the room

https://en.wikipedia.org/wiki/Carbon_nanotube_field-effect_transistor

Speaking from a nano material standpoint I question how reliable they can be for mobile applications as changing acceleration among other things can effect their readout, at least it does for similar setups I have studied.

Also while they do appear to work, the architecture and interface designs have to be remade to make them widely useful. In short one would need to redesign the computer and software to work with them, which is a huge but not insurmountable barrier that could keep them from even being widely sold. In addition to the age old issues of how to mass produce them cheaply, which can kill any new development if not solved.

Could see them in some industrial application where they can afford some high end custom solutions. Then eventually coming down to regular consumers much like how computers have come to market in the past.

With things coming to market faster and faster, who knows how long it will be. I think it will be longer as physical systems tend to still take decades as new equipment and factories still need to be built them, assuming they find a simple solutions to the issues allowing it in the first place.

>>7751031
Not the same poster but I guess he means flexible electronics.

>>7750808
Yes. Carbon has potential for ballistic transport, which is the next best thing to superconductivity. If that works out, going to ballistic transport wires and ballistic deflection transistors would be a big improvement over copper wires between silicon FETs.

Mind you, there are other plausible candidate materials to succeed silicon. We'll see what actually happens.