Google announces quantum computer error correction AI 'AlphaQubit', more accurate than existing error correction technology
In order to put quantum computers into practical use, it is essential to develop a mechanism for correcting quantum bit errors. Google DeepMind and Google Quantum AI have recently announced the error correction AI ' AlphaQubit '. AlphaQubit is said to be capable of more accurate error correction than existing technologies that use algorithms.
Learning high-accuracy error decoding for quantum processors | Nature
AlphaQubit: Google's research on quantum error correction
https://blog.google/technology/google-deepmind/alphaqubit-quantum-error-correction/
Quantum computers use 'qubits' as the basic units of information, but qubits have the problem that their state is easily affected by noise. For this reason, research and development of error correction technology that can correctly detect qubit errors is underway in order to put quantum computers into practical use.
While existing error correction technologies generally use algorithms rather than AI, AlphaQubit is an AI-based system developed using Google's AI architecture ' Transformer '. AlphaQubit is trained to correct errors in 49 quantum bits in Google's quantum processor ' Sycamore '.
Below is a graph comparing the accuracy of AlphaQubit with existing error correction techniques. AlphaQubit has 30% fewer errors than the '
In addition, AlphaQubit demonstrated superior error correction performance compared to correlated matching in simulations using large-scale quantum computers.
According to Google, while AlphaQubit is highly accurate, it has the problem that it is 'too slow for use on superconducting quantum processors that perform measurements a million times per second.'
Related Posts:
