The intense search for semiconductors that can house quantum states can be traced to the promise they hold for the future of computing. Today, even experimental single-electron transistors can only represent a digital "1" or "0" depending upon whether the charge is present or absent. However, quantum states encode bits into the wave function of the electron called its "spin" state thereby enabling a superposition of any number of bits onto a single electron.
For instance, as long as the spin of an electron is undisturbed, it can represent a 1, a 0 and any other in-between values simultaneously. When the spin of one electron interacts with that of another, the result can perform parallel computations on all the values represented in their complex waveforms.
Unfortunately, the very thing that makes quantum systems useful a superposition of values makes them even more prone to errors than classical systems. The nebulous state of quantum states or qubits can be destroyed by a wide variety of factors, all of which boil down to an inadvertent coupling to the environment resulting in decoherence of the superposition into a classical value.
To solve this problem, quantum error-correction methods were proposed as early as 1995 and first demonstrated in 1998. Since then, many groups have refined quantum error-correction encoding techniques, which basically replicate a nebulous qubit's value onto separate physical systems that are "entangled" that is, their nebulous values are synchronized over time despite different physical locations.
Entanglement enables observers to subsequently "compare" the resultant qubits after a calculation, without "observing" their nebulous values during the calculation, thereby detecting if any differences arose between the supposedly identical copies (indicating an error, which usually resets the system to try that calculation over again).
Entanglement also aids in cryptography by being able to detect eavesdropping.