A new publication from IBM has announced a step towards overcoming one of the biggest fundamental problems with quantum computers: they often just don’t work right. Quantum computers are inherently hard machines to make. From the difficulty of entangling the qubits in the first place to the errors inserted by the difficulty of keeping them entangled, at any given moment, it really does seem like quantum computers are trying not to work. In an effort to maintain entanglement, we have historically had big, shielded boxes and networking lines that protect the qubits from outside disturbance. You can also try networking with photons, creating laser-based quantum networks that exploit light’s weak interactions with air. But this inevitably is not enough. Qubits are too delicate for easy use, and the near-constant problems they encounter make quantum computing unavoidably error-prone. A few different approaches seek to address this. Collectively, they’re known as quantum error correction. A decoder is one such approach, working to find and fix the errors that plague the raw results of most quantum processes. It’s a tortured problem, though. Everything one might think to do to find and fix errors in a quantum process will disturb the entangled state of the qubits and end that process entirely. That’s true of both quantum memory (maintaining the state of qubits) and computing.

Full research : IBM’s New Decoder Could Kickstart Real-World Quantum Computing.