Google improves error correction in quantum computers

Researchers at Google demonstrated quantum computing in which the error rate decreases as the scale of error correction increases. This work means a step closer to scalable quantum error correction to enable quantum computers to achieve low enough error rates to run usable quantum algorithms. The study was recently published in Nature.

Quantum computers, like classical computers, are prone to errors caused by the “noise” (or interference) of the physical system behind them; Realizing its potential requires reducing the error rate. One method of quantum error correction is to use error correction to form a logical qubit using a set of physical qubits (units of quantum information, equivalent to bits in classical computers). This system, called surface-code logic qubits, can detect and correct errors without affecting the information, but scaling such a system means manipulating more qubits, which can introduce more logic errors. In order for logic performance to improve as the size of the code increases, the overall error correction needs to exceed the added logic error.

Hartmut Neven of Google’s “Quantum Artificial Intelligence” and colleagues demonstrated a logical qubit surface code that can reduce the error rate as the system grows in size. They built a 72-qubit superconducting quantum processor and tested it with two different surface codes: a logical qubit called distance-5 (based on 49 physical qubits) and a smaller distance-3 logical qubit (based on 17 physical qubits). Larger surface codes exhibit better logic qubit performance (2.914% logic errors per cycle) and outperform smaller surface codes (3.028% logic errors per cycle). The authors note that more research is needed to achieve the logical error rate required for efficient calculations, but this work demonstrates the basic requirements for future development. (Source: China Science News, Jinnan)


Two generations of Suzuki processors made by Google’s “Quantum AI”. Each processor has two chips glued together, one that includes qubits and one that contains the wiring between qubits and the outside world. The latest generation of Suzuki processors on the right is significantly larger than previous generations, but there are also some modifications to the qubits that improve performance. Image courtesy of Google Quantum AI

A photo of a fully assembled quantum system from Google’s “Quantum AI.” Highlighted are the diluted refrigerator where the calculations are performed, the quantum processor and quantum-limited amplifier installed on the bottom floor of the refrigerator, the various cables connecting the bottom and top layers, and the electronic control equipment of the quantum computer behind. Image courtesy of Google Quantum AI

Related paper information:

Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button