News First functional graphene semiconductor paves the path to post-silicon chips — Georgia Tech researchers' material can be used with standard chipmak...

[Admin]

Georgia Tech researchers claim to have created the world's first functional graphene semiconductor material after a decade of refinement and band gap tuning.

First functional graphene semiconductor paves the path to post-silicon chips — Georgia Tech researchers' material can be used with standard chipmak... : Read more

 

[Crazyy8]

But does it perform well? It may function, but who knows if it will outperform standard chips.

 

[bit_user]

But does it perform well? It may function, but who knows if it will outperform standard chips.

That's the idea. By improving signal propagation speed and reducing heat, my guess is they expect it can run at higher clock speeds, within the same power envelope.

Keep in mind that they only made a proof of concept to show the technology can work. Presumably, the next step would be to try and fab some fundamental logic gates on it, at which point there will hopefully be more concrete projections of the expected benefits.

 

[TCA_ChinChin]

Keep in mind that they only made a proof of concept to show the technology can work. Presumably, the next step would be to try and fab some fundamental logic gates on it, at which point there will hopefully be more concrete projections of the expected benefits.

I'm cautiously optimistic, but I imagine that even if the next steps are easily feasible, I can see another decade of research before more papers/technical breakthroughs are achieved. I feel like so many breakthroughs/research based off of graphene have taken so long to mature, it feels like a nuclear fusion scenario, where even if real tangible progress is made every year, the timeline to actual commercial use is forever in limbo.

 

[bit_user]

I feel like so many breakthroughs/research based off of graphene have taken so long to mature, it feels like a nuclear fusion scenario, where even if real tangible progress is made every year, the timeline to actual commercial use is forever in limbo.

I think that's because it's hard to form large, continuous sheets of it, and its properties are orientation-dependent. So, I think that means you pretty much have to form it in-place, rather than somehow manufacture and apply it as separate steps.

In this case, graphene wasn't up to the task, all by itself. So, there was the added dimension of how to workaround that. Then, once you have a theory, you have to figure out how to fabricate the hybrid material.

 

[LuxZg]

"Tianjin University in China" -> oh no, they have already stolen the designs! ;-)

"Walter de Heer, Regents’ Professor of physics at Georgia Tech, led a team of researchers based in Atlanta, Georgia, and Tianjin, China, to produce a graphene semiconductor that is compatible with conventional microelectronics processing methods — a necessity for any viable alternative to silicon."

"Over the next decade, they persisted in perfecting the material at Georgia Tech and later in collaboration with colleagues at the Tianjin International Center for Nanoparticles and Nanosystems at Tianjin University in China. De Heer founded the center in 2014 with Lei Ma, the center’s director and a co-author of the paper."

Need to save these quotes for discussions another 10 years down the line when everyone start spreading FUD...

Anyway, back to news reading. This is real interesting, but same as bunch of other tech will take a lot more time to actually show up in everyday devices like PCs, phones, cars... Hope I'm still alive when it happens! (not that it will take THAT long, but I'm also not young anymore)

 

[vehekos]

Graphene cannot preserve Moore's law, because it cannot be miniaturized much. It needs a lot of atoms just to be graphene.

 

[bit_user]

Graphene cannot preserve Moore's law, because it cannot be miniaturized much. It needs a lot of atoms just to be graphene.

Like how many? Is 10 enough?

Even if they do turn out to be bigger, that can probably be worked around by layering or stacking, since they run cooler.

I think it's a mistake to second-guess them, especially given how much focus this area has gotten. If it really didn't have the potential for miniaturization to anywhere near leading edge nodes, I doubt they've have continued investigating it.

 

[DavidC1]

Yea, such "Wonder" material which will cause serious health issues down the road since practically every area is trying to use nano tech and graphene including and not limited to waterproof coating, antibacterial lining, and paint. The material breaks off similar to flint meaning sharp and long and the micro structures will easily be carried off into the wind into your lungs and other systems.

Use it for CPUs and forget everything else please.

 

[bit_user]

Yea, such "Wonder" material which will cause serious health issues down the road since practically every area is trying to use nano tech and graphene including and not limited to waterproof coating, antibacterial lining, and paint.

That has nothing to do with this. A CPU is not going to leak out nanoparticles of anything.

The material breaks off similar to flint meaning sharp and long and the micro structures will easily be carried off into the wind into your lungs and other systems.

Well, it's ultimately just carbon. So, it will eventually react with other elements in the environment, as opposed to staying like that.

Sure, the health & environmental implications of putting nanoparticles of anything into various products needs to be understood and managed, but I wouldn't single out graphene, here.

 

[The Beav]

Graphene cannot preserve Moore's law, because it cannot be miniaturized much. It needs a lot of atoms just to be graphene.

Moore's law is already slowing down with traditional silicon, we're getting close to single atom separation now with 2nm, and there's also a very large interference issue that's cropped up and stopping further miniaturization. So Moore's law will essentially stop, or slow down, unless other avenues of miniaturization are explored. If the atoms in graphene are smaller than those in silicon this could be a huge advantage in miniaturization.

 

[bit_user]

If the atoms in graphene are smaller than those in silicon this could be a huge advantage in miniaturization.

Graphene, itself, is just a planar arrangement of Carbon atoms. However, I have no idea about how structures made from "epitaxial graphene bonded to silicon carbide" would scale, in relation to conventional transistors. If there were a major size benefit, I assume that probably would've been mentioned.

BTW, Moore's Law doesn't actually deal with speed, but rather density. So, in the purest sense, this development might not reinvigorate it, but would still hopefully pave the way for continued performance & efficiency improvements.

 

[LuxZg]

BTW, Moore's Law doesn't actually deal with speed, but rather density. So, in the purest sense, this development might not reinvigorate it, but would still hopefully pave the way for continued performance & efficiency improvements.

This is the correct way of thinking. Moore's "law" isn't law, it was just some rough prediction based on statistics and past experiences. And it held pretty good for 60-70 years. But we don't care about number of atoms, or spacing, or amount of transistors. All we care is price, performance, and power, and then the values extrapolated from them like price/performance and performance/power. If these graphene based materials can give us more performance for less power and at the same time keep pricing relatively "normal" (which promises of using existing toolchain suggests), then it's good enough that it will matter. It's just a long road ahead. They can just be too slow to matter by the time they're ready, new breakthrough in silicon or other materials might happen, or whatever. So it's mostly tech to keep an eye on from time to time. And it sounds as if it might give our existing transistor based computing another stretch of life.

 

[yur0k]

So, China already has the tech before it's been released, nice!