IBM uses sellotape to improve graphene based chips
Stick that in your nanopipe and smoke it
ACCORDING TO TECHNOLOGY REVIEW, boffins from IBM have developed a way to improve transistor performance and reduce a device’s level of electrical noise by simply stacking two layers of graphene on top of each other.
Their findings, published in the current issue of the journal Nano Letters, show that piling up layers of graphene could make chips faster, smaller, and less power consuming than those based on silicon.
Graphene, which on an atomic scale resembles chicken wire, is made up of a single layer of SP2 bonded carbon atoms, densely packed into a honeycomb crystal lattice. It is ideal for use in integrated circuits because of its high carrier mobility. But despite the fact that it can carry higher currents at faster rates than silicon (some estimates put enhanced electron mobility anywhere from 50 to 500 times faster), noise is still the number one problem for a graphene chip based device.
Because graphene chips tend to be even tinier and more compact than silicon ones, the miniscule currents flowing through them are more sensitive to outside influences, resulting in significantly greater noise.
But Biggish Blue boffins reckon that by using sellotape pieces to remove layers of graphite and by putting them one on top of the other, they can reduce the electrical noise of a device by a factor of 10.
But don’t get too excited just yet. It will probably take quite a while before the new and improved graphene transistors will be able to be manufactured en masse. µ
L'Inq
Technology Review
Comments
.
Charges rush forwards together holding hands with no eyes for the maddening crowd, how romantic.Cool, now we can go from "gigahertz don't matter" to "nanometers don't matter."
50 to 500 times as fast? Okay, let's take the worst case scenario and assume 50 times faster is correct. So, a chip of the same size could be 50 times faster.I have to ask the question, if it's 50 times faster, does that mean that the chip can be 50 times bigger? Or could it be a little over 7 times bigger(square root of 50) and still get the same speed? Or is it something else entirely?
No.
No, 50x faster electron mobility does *not* make a transistor switch on/off 50x faster, and it most certainly does not make a CPU 50x faster.You might as well expect a faster burning fuel to make your car run faster.