Google Quantum AI Adds Neutral Atom Computing to Superconducting Roadmap

Google Quantum AI is betting on two horses in the race toward commercially viable quantum computers. The tech giant announced it’s expanding beyond superconducting qubits to build neutral atom quantum systems—a move that could accelerate its timeline for hitting key milestones before decade’s end.

The dual-modality approach exploits what Google calls “complementary strengths.” Superconducting qubits excel at circuit depth, running millions of gate cycles at microsecond speeds. Neutral atoms scale better in qubit count, with arrays reaching roughly 10,000 qubits and offering flexible connectivity that enables more efficient error-correcting codes. The tradeoff? Neutral atom cycles run in milliseconds, orders of magnitude slower.

Boulder Becomes Google’s Atom Hub

To lead the experimental push, Google recruited Dr. Adam Kaufman from JILA, the joint institute between CU Boulder and NIST. He’ll build a neutral atoms hardware team in Boulder while maintaining his JILA Fellowship and CU faculty position—a setup that keeps Google embedded in one of the world’s premier AMO (Atomic, Molecular and Optical) physics ecosystems.

“I am thrilled to join Google’s world-leading program in quantum computing, and to expand that leadership to a new and highly promising platform of neutral atoms,” Kaufman said.

NIST’s Physical Measurement Laboratory Director James Kushmerick acknowledged the talent drain but framed it positively: “While this is a loss for NIST, it is a gain for the quantum ecosystem in Boulder and the U.S. quantum industry broadly.”

Building on Willow’s Momentum

The expansion comes after Google’s superconducting program hit a major milestone in December 2024. The 105-qubit Willow chip became the first processor to operate below the quantum error correction threshold—meaning errors actually decreased as the system scaled up, rather than compounding. That’s the fundamental requirement for fault-tolerant quantum computing.

Willow also demonstrated what Google claims is verifiable quantum advantage, completing a random circuit sampling benchmark in under five minutes that would theoretically take the Frontier supercomputer 10 septillion years.

Google’s neutral atom program will focus on three pillars: adapting error correction protocols for atom array connectivity, using Google’s compute resources for hardware simulation and design optimization, and building application-scale hardware with fault-tolerant performance.

QuEra Connection Remains

The announcement notes Google will continue collaborating with QuEra, its portfolio company that pioneered foundational neutral atom methods. That relationship gives Google additional exposure to the technology while building internal capabilities.

For the broader quantum computing sector, Google’s dual-track investment signals growing confidence that multiple qubit technologies may prove commercially viable—each potentially suited to different problem types. The company reiterated its belief that “commercially relevant quantum computers based on superconducting technology will become available by the end of this decade,” with neutral atoms potentially offering complementary capabilities.

The next major test for neutral atoms: demonstrating deep circuits with many computational cycles. For superconducting systems, it’s scaling architectures to tens of thousands of qubits. Google’s now racing on both fronts.

Image source: Shutterstock