Google says it has made a breakthrough in quantum computing that reduces errors

A major challenge is that quantum computers can store or manipulate information incorrectly, preventing them from running algorithms long enough to be useful. The new research from Google Quantum AI and its academic collaborators shows that they can actually add components to reduce these errors. Previously, adding more components to a quantum computer often led to more errors due to limitations in the technology. Ultimately, the work reinforces the idea that error correction is a viable strategy for building a useful quantum computer. Some critics have questioned whether it was an effective approach, said Duke University physicist Kenneth Brown, who was not involved in the research.

“These kinds of error corrections really work, and I think they’re only going to get better,” Google team member Michael Newman wrote of X. (Google, which posted the study to the arXiv preprint server in August, declined to comment on the data for this story.)

Quantum computers encode data using objects that behave according to the principles of quantum mechanics. In particular, they store information not only as 1s and 0s, as a conventional computer does, but also in “superpositions” of 1 And 0Storing information in the form of these superpositions and manipulating their value using quantum interactions such as entanglement (a way in which particles can be connected to each other even over large distances) makes entirely new types of algorithms possible.

In practice, however, quantum computer developers have discovered that errors quickly creep in because the components are so sensitive. A quantum computer represents 1, 0or a superposition by placing one of its components in a certain physical state, and it is too easy to accidentally change those states. A component then ends up in a physical state that does not correspond to the information it is supposed to represent. These errors accumulate over time, meaning that the quantum computer cannot provide accurate answers for long algorithms without error correction.