Google Eliminates Errors in AI
There is an AI based decoder that identifies quantum computing errors. It is being introduced from the AI researchers at Google Deep Mind, working with a team of quantum researchers at Google Quantum AI.
The paper was published in the journal Nature. The team used machine learning to help find qubit errors more effectively than previous ways. Nadia Haider is with Delft University of Technology's Quantum Computing Division of QuTech. The team published a News and Views piece in the same journal outlining the work done at Google.
The main problem with developing a useful quantum computer error correction. Qubits are fragile, and their quality is not great, often resulting in errors. The team at Google have developed an AI based decoder to identify these errors.
Google's Artificial Intelligence division has worked on a quantum computer called Sycamore. It creates single logical qubits that use multiple hardware qubits. These are used to run programs while carrying out error correction. The new program has been named Alpha Qubit. The new AI decoder is known to be a type of a deep learning neural network. Using their Sycamore computer, researchers trained it to recognize errors. They ran 49 qubits through a quantum simulation.The two systems created hundreds of millions of quantum error examples. They then used Alpha Qubit to identify the errors and all of them were corrected.
The testing resulted in a 6% improvement in error correction, and a 30% improvement when running less accurate but faster tests. They also tested it using 241 qubits and found great results. The findings show machine learning is the solution to error correction of quantum computing, as it allows for concentration on other problems yet to surface.

Comments
Post a Comment