Quantum Computers: Google Now Also Researching Neutral Atoms

Google's research division, Google Quantum AI, is now exploring quantum computers based on neutral atoms in addition to superconducting qubits. This was announced by Hartmut Neven yesterday on the company's blog.

For decades, the company has prominently relied on superconducting qubits. In 2019, the team demonstrated quantum supremacy for the first time with its quantum chip. Last year, (former) Google researchers John Martinis and Michel Devoret, along with John Clarke, won a Nobel Prize in Physics for the fundamental principles that made superconducting quantum chips possible.

Complementary Strength

The team will not stop its work on superconducting qubits. “We are confident that commercially viable quantum computers based on superconducting technology will be available by the end of this decade,” Neven writes on the blog.

Instead, the goal is to leverage the complementary strengths of both architectures. Superconducting qubits consist of tiny circuits that exhibit quantum properties at very low temperatures near absolute zero, which can be used to develop qubits.

On superconducting chips, researchers have already been able to execute deep circuits with millions of quantum gates. This is achieved within microseconds. However, the circuits are rigid, the qubits are permanently implemented on them, and interconnecting the qubits is therefore a challenge. It is also difficult to increase the number of qubits to tens of thousands, as errors accumulate and quantum information becomes noisy.

Atoms are the fundamental building blocks of nature and are identical by nature. Two isolated states of atoms serve as qubits, and atoms have the advantage of being more robust against errors than superconducting qubits. It is also easier to arrange atoms precisely in space and to shift a calculation during runtime. Thus, research groups have already been able to combine tens of thousands of atoms into so-called arrays.

However, quantum gates with atoms are slower; they require milliseconds, making them about a thousand times slower than with superconducting qubits. This is largely compensated for by higher stability; nevertheless, no deep circuits with neutral atom qubits have been demonstrated yet.

The combination of the two platforms could therefore enable a step forward. “By investing in both approaches, we can achieve our goal faster,” Neven writes. “By advancing both areas, we foster mutual exchange between research and technology and can thus offer access to versatile platforms tailored to different problem sets.”

Building Blocks of Research with Neutral Atoms at Google

The team is focusing on several aspects in its research with neutral atoms. The first is quantum error correction. While the team has already shown significant progress in correcting quantum errors with superconducting qubits, this step for neutral atoms is still pending. The higher connectivity could be an advantage here.

Furthermore, the team intends to further develop the hardware. The goal is to control atomic qubits in large numbers and reduce errors. The team wants to support these efforts with computer-aided simulations and modeling.

Adam Kaufman will lead the newly created hardware team for neutral atoms. The physicist previously researched atoms at JILA, a research institute of the University of Colorado (CU) Boulder and NIST. However, he will continue to serve as a JILA Fellow and a member of the CU Boulder faculty, belonging to the CU Boulder Department of Physics. With this, Google aims to underscore its proximity to the quantum research landscape in the USA.

In addition, Google is continuing its collaboration with QuEra. QuEra is a spin-off of Harvard University and MIT and is currently considered one of the most promising startups in the field of neutral atom quantum computers. Google Quantum AI already invested in the startup at the end of 2024.

German research groups are also investigating quantum computers based on neutral atoms. These include the DLR Quantum Computing Initiative and the startup planqc from Garching near Munich. Neutral atoms are considered one of the most promising architectures for building powerful quantum computers.

(spa)