Oxford Researchers Demonstrate Fast, 99.8% Fidelity Two-Qubit Gate Using Simplified Circuit Design

Insider Brief:

• Oxford researchers demonstrated a 25-nanosecond controlled-Z gate with 99.8% fidelity, combining high speed and accuracy in a simplified superconducting circuit.

• The design pairs two qubits with opposite anharmonicities, eliminating the need for tunable couplers and reducing circuit complexity.

• By accessing a third quantum state during the gate operation, the team achieved a theoretical √2 speed-up, confirmed experimentally.

• The result advances fault-tolerant quantum computing by improving raw gate performance without relying heavily on error correction or added hardware.

• Image Credit: Oxford Quantum Circuits

Speed and fidelity are two of the most meaningful parameters for quantum gates. The faster the gate, the less time for decoherence to erode fidelity. The higher the fidelity, the less a system has to rely on quantum error correction, which, while powerful, comes with considerable overhead. A new result from the University of Oxford takes a direct shot at both.

The Sweet Spot: Fast, High-Fidelity Gates

As announced in a press release from the University of Oxford and presented at the American Physical Society’s Global Physics Summit on March 21, researchers have demonstrated a controlled-Z gate operating at 99.8% fidelity in just 25 nanoseconds. The CZ gate is a fundamental two-qubit operation used to create entanglement, and improving its performance is essential for building large-scale quantum computers.

The team, led by Dr. Peter Leek and presented by Dr. Simone Fasciati, achieved this using a modified coaxmon circuit architecture—a three-dimensional superconducting qubit design known for its simplicity. Their new approach pairs a standard transmon qubit with an inductively shunted transmon, a specially designed variation that behaves differently at higher energy levels. These two qubits have opposite “anharmonicities,” meaning their energy levels are spaced in different ways, which enables both faster gate speeds and suppression of a common error.

That error, known as ZZ crosstalk, occurs when two qubits unintentionally influence each other while idle, leading to unwanted shifts in their states. Many existing systems address this by adding tunable couplers—extra hardware that isolates qubits—but doing so increases circuit complexity. According to the release, the Oxford team eliminates the need for such couplers through the natural interaction of the two differently tuned qubits. This allows them to maintain a fixed coupling, which has the added benefit of simplifying the architecture.

The speed-up comes from using not just two, but three quantum states during the gate operation. Typically, CZ gates rely on an interaction within a two-state subspace. But theoretical work suggested that accessing a third state could boost speed by a factor of √2. The Oxford team’s setup makes that possible, and their results confirm the effect experimentally.

“It’s one of the fastest 2Q gates ever benchmarked, has a very competitive gate fidelity, and one of the simplest circuit designs,” said Dr. Peter Leek in the release. “That combination bodes extremely well for scaling up to perform fast and useful quantum computing.”

The Role of Gate Performance in Fault-Tolerant Quantum Systems

The significance of the result lies in hitting both speed and fidelity at once—something rarely achieved together. Quantum error correction, which compensates for hardware imperfections, generally assumes a baseline fidelity of 99% or higher. Once that threshold is passed, remaining errors can be handled algorithmically. But QEC is expensive: most approaches require encoding a single “logical” qubit into dozens or hundreds of physical qubits. So every bit of raw hardware improvement we can build in reduces this burden. In this case, 99.8% fidelity puts the gate well into fault-tolerant territory.

Speed matters, too. Qubits are inherently fragile; information begins leaking the moment a gate starts. Performing an operation in just 25 nanoseconds limits the opportunity for that loss, which effectively preserves the fidelity even before correction kicks in.

Dr. Fasciati described the concept behind the design as “simple and elegant: to make more efficient use of the computational resources already available in a two-qubit system.” He also noted that the principle—getting more out of existing circuits—could potentially be applied beyond this specific case.

A patent has been filed on the technique for achieving opposite anharmonicities. Oxford Quantum Circuits, where Leek now serves as CSO, is already using coaxmon-based technology in commercial settings. Previous work from the same lab on a different type of two-qubit gate, the cross-resonance gate, was later scaled up and optimized by OQC in their systems. According to the release, the current work could follow a similar path, moving quickly from lab demonstration to industry use.

Performance Without Complexity

While this result doesn’t solve the broader scaling challenge in superconducting quantum systems—such as routing control wires across increasingly crowded chips—it shows that better performance doesn’t have to come at the cost of added complexity. By refining the underlying physics and improving the performance of fundamental operations, the Oxford team’s work offers a clean proof-of-concept for practical, high-speed entanglement in commercial systems. With each gate that becomes faster, cleaner, and simpler, building useful quantum systems becomes a little more possible.