Bharath Kannan, 27, found a new way to make quantum computers more powerful by reducing the error rates of their fundamental computing units, known as qubits. Now the company he leads, Atlantic Quantum, is working to build quantum computers that he hopes will advance the fields of encryption, materials science, and machine learning.
What makes quantum computers so powerful is that qubits can exist in multiple states at the same time, allowing for new types of calculations. But qubits are highly unstable and error prone, so companies trying to build quantum computers powerful enough to tackle real-world problems end up at an impasse. Each additional qubit intended for running a calculation may require thousands of additional qubits devoted solely to canceling out its errors. Unfortunately, the problem grows exponentially rather than linearly, so you can’t easily solve it by brute force.
Kannan and Atlantic Quantum have taken a very different approach: building a better qubit. Their so-called fluxonium qubits operate at a much lower frequency than the type used by many other companies, called transmon qubits. At this lower frequency, there is less overlap between signals, which means less interference and fewer errors.
Researchers had known about fluxonium qubits for roughly a decade. But their advantages went unexploited because they were more complex to design and build, and more difficult to control.
Kannan solved those challenges by developing new circuits for building and controlling low-frequency qubits. As an added benefit, he demonstrated that fluxonium-based quantum computers, because they operate more slowly without sacrificing processing speed, can be controlled with simpler hardware than other types of quantum computers require.
