Quantum Computers and Signal Noise

Quantum computers represented a possible "next-step" for the tried-and-true conventional binary computing used in much of the world. While conventional computers represented data using binary on and off switches (each "bit" of data existing in one of two states), quantum computers takes advantage of the superposition principle in quantum physics that allow it to represent data in one of three possible states. This includes the on and off switches of a conventional computer and a distinct state where a given "qubit" (analog to the conventional "bit") can be both on and off. This can be particularly useful for computationally difficult problems, where testing every possible solution would take the conventional computer a couple million years. The quantum computer can skip a lot of these tests, almost as if it were testing multiple solutions at any one step. 

Unfortunately, signal noise has a tendency to introduce errors during computation. This can actually alter and corrupt the state that a bit (and by extention, qubit) is in, thereby corrupting whatever representated data that bit is a part of. A weakness of quantum computing is its obvious reliance on a basic principle of quantum physics, which make redundancy and repetition (a simple matter in conventional computing) impossible as a method of error correction. Researchers have also found that their current method of error avoidance (as opposed to the reactive error correcting) cannot be used on its own, dashing any hopes of using that method exclusively in future implementations of the quantum computer.

Then again, whatever that is determined to not work only narrows down the possible choices of what will work, so research continues to advance.

Quantum Computers Will Require Complex Software To Manage Errors, ScienceDaily