Graphene said to replace current level PCBs and electronics

http://web.mit.edu/newsoffice/2009/graphene-palacios-0319.html

Blast from the past, but this is the thread I found when I searched to see if it ever got posted.

New material could lead to faster chips

Graphene may solve communications speed limit
..but the new graphene technology could lead to practical systems in the 500 to 1,000 gigahertz range.
But Kong has been developing a method for growing entire wafers of graphene directly, which could make the material practical for electronics.

Doesn't this stuff have the same advantages of the M.I.A. diamond processors since they are both carbon?
 
Last edited by a moderator:
Standard silicon transistors already switch in the hundreds of GHz range, so a single clock multiplier topping out at maybe 2-3 times faster, while notable, is not sufficient to lead to chips hundreds of of times faster.
 
I thought the difference here was that whole systems could run that fast, not just exhibition single transistor demos.

Frequency multipliers are widely used in radio communications and other applications. But existing systems require multiple components, produce "noisy" signals that require filtering and consume large power, whereas the new graphene system has just a single transistor and produces, in a highly efficient manner, a clean output that needs no filtering.

Which would result in less heat, no?
 
Clock signal generation isn't what's holding back current processors, and 1 THz would only allow a signal to move less than a third of a millimeter, assuming it had no interaction with other logic and was moving at speed of light in a vacuum.

Whatever benefits graphene may offer, an unusable clock cycle time is not going to be the biggest benefit.
 
It doesn't seem to be a very sturdy material, so I can hardly imagine this replacing anything soon, especially where more power (in electrical sense) is needed.
 
You can already access THz part of spectrum w/o using graphene. These things are forever 5-10 years away from industrial applications. Though graphene is easily the hottest material of 2009. Yet before you dream up a 1THz Core i23720 made of graphene, care to think of what happened to CNT's.
 
It doesn't seem to be a very sturdy material, so I can hardly imagine this replacing anything soon, especially where more power (in electrical sense) is needed.

Strength is irrelevant since it's sitting on a substrate. If your building bulk stuff out of graphene it's up to 200 times stronger than steel on a per weight basis.

You'd probably use carbon nanotubes instead in the near future because it's much further along in commercialization. Big CNT sheets(4 feet by 8 feet) are already being made as well as wires with unprecedented performance compared to any other material(even though these CNTs are performing far below their full potential so far). It's going to get used in aircraft skins and aircraft wiring, satellites and so forth first; when it comes down in price you're going to see it popping up just about everywhere aluminium and copper wiring is found or exceptional strength is needed(electric motors, power lines, radio antennas, kayaks, wear resistant safety clothing...).
 
If you want speed, get rid of the clock altogether. And if you want even more speed, have every part execute at it's own leisure.

It can be done, it's simply a totally different ballgame. But I think it's the only one with a lot of potential, as far as common chip technology is concerned.

Another potentially interesting road is nanomachinery. Go back to pure mechanics, forget about electricity and use molecules that push and pull each other.

Quantum computers are like fusion plants and artificial intelligence: ever 50 years in the future.
 
You can already access THz part of spectrum w/o using graphene. These things are forever 5-10 years away from industrial applications. Though graphene is easily the hottest material of 2009. Yet before you dream up a 1THz Core i23720 made of graphene, care to think of what happened to CNT's.

Aren't CNT getting closer and closer to commercial use?

Graphene, from the article, seems to have not such a large set back from testing to actual commercial production.

Strength is irrelevant since it's sitting on a substrate. If your building bulk stuff out of graphene it's up to 200 times stronger than steel on a per weight basis.
It's going to get used in aircraft skins and aircraft wiring, satellites and so forth first; when it comes down in price you're going to see it popping up just about everywhere aluminium and copper wiring is found or exceptional strength is needed(electric motors, power lines, radio antennas, kayaks, wear resistant safety clothing...).

Does this mean I won't have to use giant wires for my subwoofers anymore??
 
From http://poplab.ece.illinois.edu/pdfs/Grosse-GrapheneContactSJEM-nnano11.pdf via http://www.nyteknik.se/nyheter/innovation/forskning_utveckling/article3148442.ece via http://www.tekniikkatalous.fi/tk/nanotekniikka/article608895.ece

Researchers have been for the first time able to measure active graphene transistor temperature, and noticed that when electricity goes through it, the other side cools while the other heats, acting like a peltier - they suspect that the cooling power might have been greater than the amount of heat produced by electricity, which could lead us to have self-cooling transistors in our hands apparently.
 
Any idea where the energy goes? It's not like they broke physics laws by actually destroying it.

Nope, I haven't even read through the PDF myself (and wouldn't prolly understand half of it)
 
Any idea where the energy goes? It's not like they broke physics laws by actually destroying it.

Does it matter? There's no need to break laws of physics, it's simply heat transfer. A refrigerator is similarly capable of transporting more heat from one place to another than what it produces itself.
 
Interesting but I see no tangibe benefits from this as you'd still need some kind of heatsink/fan just like in a Peltier heat pump.
 
Interesting but I see no tangibe benefits from this as you'd still need some kind of heatsink/fan just like in a Peltier heat pump.

in the same vein there will always be the need to remove heat from a computer case, remove heat from a whole server room, etc. so this new feature deals only with the bottom of the cooling chain, the chip itself.

if it leads to a higher range of chips that work with no heatsinks, or thinner heatsinks that still is useful.
 
Back
Top