Willow: What Google's landmark error-correcting chip means for the world of quantum computing

NEW DELHI: Google has just announced a new quantum chip, Willow, and the breakthrough needs to be read right in order to understand its implications.

- Kabir Firaque

Yes, it is an exciting fact that Willow took under 5 minutes to perform a benchmark computation for which an existing supercomputer would need 10 septillion years, which is 1 followed by 25 zeros. The real achievement, however, is not just the computation, but also about a long-standing challenge that was conquered in getting Willow to perform it.

The keyword is "error correction below threshold", something that headlines Google's paper in Nature on the chip. This effectively means that a higher number of qubits (short for quantum bit, quantum computing's equivalent of the bit) increases error correction exponentially. The concept already existed in theory, and Google's demonstration holds the promise of scaling it up to quantum computers of the future.

What the achievement foretells, however, is best understood if one looks first at the challenge that preceded it.

What it means

Qubits, the basic units of quantum information, are prone to errors caused by interactions with the environment, with each other, and random fluctuations. These can cause qubits to lose their properties and hamper the quest to build mass-use quantum computers that can ultimately replace classical computers. The immediate objective, therefore, has been to find ways to correct those errors.

The idea is to incorporate error correction codes in "logical qubits", each created across a number of "physical qubits". When we talk of a physical qubit, we are referring to a piece of physical hardware in the form of an electron or a photon. A logical qubit is an abstraction rather than a single physical object, a system that is implemented using the physical qubits that it is built across. As such, it works in a way that is more complex than any of these physical qubits does.