Quantum gates set speed record at 6.8 nanoseconds with over 99.98% accuracy

Researchers from Jülich, Aachen, and Cologne have achieved a new speed record for quantum operations, performing a quantum gate on a qubit in just 6.8 nanoseconds with more than 99.98% accuracy. This speed significantly surpasses typical operations in superconducting quantum computers, which usually take several tens of nanoseconds. The findings were published in Nature Communications on October 27, 2026.

The 6.8 nanosecond duration is comparable to the time light travels approximately two meters in a vacuum. This switching speed is a critical factor in quantum computing because qubits are highly sensitive and can only maintain their quantum states for a limited time. Faster individual operations allow for more complex algorithms and efficient resource-intensive error correction within this brief window.

However, the use of shorter, stronger microwave pulses to switch qubits increases the risk of unwanted excitation. Superconducting qubits have more than two energy states, and strong pulses can unintentionally push them to higher energy levels through multiphoton transitions. These transitions, previously overlooked, are particularly problematic as they can persist across multiple computational steps and impair quantum error correction.

A collaborative team including Yuan Gao and Professor Rami Barends from the Peter Grünberg Institute for Functional Quantum Systems (PGI-13), alongside José da Costa Jesus, Dr. Francisco Cárdenas-López, and Professor Felix Motzoi from Quantum Control (PGI-8) at Forschungszentrum Jülich, developed a novel method to suppress these unwanted transitions. Their R2D method precisely shapes control pulses, acting like a frequency filter to suppress certain frequency components, ensuring the desired computational operation while minimizing excitation to higher energy states.

Building upon the established DRAG method, which has been a standard correction technique for over a decade, the new R2D method reduces the probability of unwanted multiphoton transitions by more than 20-fold. The method was tested on a single superconducting transmon qubit fabricated at the Helmholtz Nano Facility at Forschungszentrum Jülich.

Two fundamental quantum operations were performed, each lasting 6.8 nanoseconds, achieving measured accuracies of 99.981% and 99.989% respectively. The probability of the qubit unintentionally transitioning to a higher energy state remained below 0.002% per operation. While these results set a new benchmark, further experiments are necessary to determine the method's effectiveness in larger quantum processors and during operations involving two qubits.

What to watch: Future experiments on R2D method's scalability and multi-qubit performance.

Editor's note: The article effectively summarizes the technical breakthrough, team members, and the R2D method while maintaining accuracy.