Researchers at RIKEN and the University of Tokyo claim that the there may be a way to build semiconductors that do not leak any power at all.
While it is key in today's semiconductor research to contain leaking current, researchers said that a switch to a more exotic semiconductor, a magnetic topological insulator, can achieve the goal of ultimate power efficiency, at least as far as power supply is concerned.
Their work is based on a finding from 1980, the quantum Hall effect, which can facilitate dissipationless electricity channels, if magnets 100,000 times stronger than the earth's magnetic field help channel the stream of electrons. Instead of relying on such massive magnets, the scientists discovered that they can use a material's own magnetic properties to achieve the same result. The researchers say the Dirac fermions of the magnetic topological insulator interact with magnetic ions, transporting current and behaving as if they have zero mass and enabling the transfer of current without leaking current. Prototype devices that prove the theory have been created by the team.
Of course, it's not a free lunch, either. The transistors only work in "cryogenic" conditions, but there is hope that the temperature requirement can be softened to take these transistors into the commercial market.
its different, at least from what i remember... cryogenic is cold, but for super conductors, they use liquid helium... or i'm thinking of something else.... which i may be.
Yes this is more or less correct. However, to get specific the point where most scientists agree cryogenics begins and refrigeration ends is at 123 K. Without more information it can be assumed that "cryogenic conditions" in this case means anything below ~123 K. As for superconducting materials, most of them super-conduct below 50 K and require use of liquid neon, liquid hydrogen, or liquid helium.
Leakage current is one of the major show-stoppers when it comes to further reducing the size of a transistor as it destroys the properties of a semiconductor (it behaves like a conductor). Heat is simply created by the resistance of the circuit. Imagine electrons (electric current) hitting other electrons (in the material) and speeding them up -> heat.
So an ideal semiconductor would be superconductive (no loss due to resistance) and have no leakage current. Usually those two don't go well together: superconductive == loss free movement of charge carriers, no leakage current = perfect control of said charge carriers. In addition most superconductive materials fail in high magnetic fields.
For more information ask the internet
Well, i'm sure they can do that with bulldozer chips. I mean, you have to bathe them in liquid nitrogen anyway to OC.
I'm just trolling a bit, don't kill me.
The heat from CPUs comes from current traveling through a resistance. Plain-and-simple. That resistance is in the wire traces, poly-Si traces / gates, and transistor channels on the chip. The highest resistance is in the channels (especially the p-transistor channels), then the poly-Si, then the metal traces. The current that travels through the resistance is either switching current or leakage current. These guys are proposing to reduce the leakage current, but that still won't solve the problem of switching current, which is tied to V^2, C, F, and a couple of other factors that depend on the nature and purpose of the IC. To solve heat-induced switching current, you need to replace the Poly-Si and channel materials with lower-resistance materials that will also enable you to lower V, or else you need to make C lower by generally making everything smaller. Only so far you can take that with Silicon.
Show off
They are CHANNELING electrons.The breakthrough is than instead of using giant magnets to avoid dispersion;they use quantum hall effect.