D-Wave just published the best quantum error correction result yet, and nobody saw it coming

[Transformer Elk]

D-Wave has spent years as quantum computing's punchline, a company selling quantum annealers that most physicists never really considered proper quantum computers, while IBM, Google, and Microsoft got the actual attention building gate model machines. On August 5, D-Wave published a peer reviewed paper in Nature demonstrating a two qubit entangling gate with 99.9 percent fidelity using something called dual rail erasure qubits, with bit flip errors suppressed to roughly one in a million operations.

The actual problem this targets is the industry's biggest bottleneck, error correction overhead. IBM's current approach needs roughly 1000 physical qubits bundled together to create one reliable logical qubit, and Google's Willow chip demonstrated below threshold error correction back in December 2024 but still with enormous overhead attached. D-Wave is claiming a way to cut that overhead by a factor of ten.

The actual trick, designed by Rob Schoelkopf, who co-invented the transmon qubit most of the industry already runs on, is genuinely elegant. Instead of storing quantum information in a single circuit, the dual rail design splits it across two microwave cavities sharing exactly one photon, computational states being photon in the left cavity or photon in the right cavity. If the photon leaks away, which is the most common failure mode in superconducting systems, both cavities end up empty, and empty is something the system can actually detect and flag immediately rather than silently corrupting a calculation.

This result specifically validates D-Wave's 550 million dollar acquisition of Quantum Circuits Inc. And it's a notable milestone given D-Wave is now apparently the only quantum computing company running both annealing and gate model architectures simultaneously, hedging its bets while most competitors commit to a single specific approach.

Worth flagging that the source covering this (Qlosophy) writes with a pretty dramatic. Advocacy-style voice throughout its coverage, so I've stuck to the parts that trace back to the actual Nature paper and D-Wave's own technical claims rather than the site's own framing

The quantum company everyone laughed at just published the best error result in the worldwww.qlosophy.com