Princeton's qubits stay stable 15 times longer than Google's

The core problem with quantum computers has always been fragility. Qubits—the quantum equivalent of classical bits—lose their information almost instantly when exposed to environmental interference, a phenomenon called decoherence. It's like trying to do math while someone keeps erasing your whiteboard. The faster the erasure, the fewer calculations you can complete before errors pile up.

Google and IBM's superconducting qubits, the industry standard, hold stable for less than 100 microseconds. That's not a typo—we're talking about one ten-thousandth of a second. A team at Princeton University just demonstrated qubits that stay stable for 1.6 milliseconds. That's 15 times longer, and three times better than anything previously achieved in a lab.

"This advance brings quantum computing out of the realm of merely possible and into the realm of practical," said Andrew Houck, who co-led the research. The shift matters because stability buys time—time for the quantum computer to actually perform useful operations before decoherence corrupts the result.

The Material Problem Nobody Solved

The breakthrough came from a materials insight that had been lurking in plain sight. Superconducting qubits are traditionally made from aluminum, but aluminum surfaces are riddled with tiny defects that absorb energy as it travels through the circuit, introducing errors. Princeton's team switched to tantalum, a metal with far fewer surface defects.

But there was a second problem hiding underneath. Earlier experiments with tantalum qubits, dating back to 2021, were built on sapphire substrates—and the sapphire itself was leaking energy. The researchers replaced it with silicon, the same material that powers the entire computing industry. Silicon can be manufactured to extremely high purity, and because it's already mass-produced globally, the new design could theoretically scale far more easily than previous lab-only prototypes.

The challenge was creating a clean enough interface between tantalum and silicon while maintaining superconductivity. The team developed a new fabrication process to solve it—a process that, importantly, is compatible with existing semiconductor manufacturing equipment.

Retrofitting the Existing Fleet

To validate the design, the researchers built a working quantum chip with six of the new qubits. The design is similar enough to current Google and IBM qubits that it could slot directly into existing quantum processors without a complete redesign. That compatibility matters. If companies can upgrade their current machines with more stable qubits, they suddenly have more time to run calculations before errors take over—potentially solving larger problems without waiting for entirely new hardware.

The path from lab to commercial production is rarely straightforward. Material science breakthroughs often look better on the bench than they perform in a foundry, and the semiconductor industry moves cautiously. Whether Google, IBM, or others will actually adopt this architecture remains an open question.

But the research has tackled one of the fundamental barriers holding back superconducting quantum computing. Stability isn't the only problem quantum computers face—error correction, scaling, and coherent operation times all matter. Yet solving decoherence by a factor of 15 is the kind of progress that shifts what's possible next.