Fermilab study traces why some superconducting qubits perform better than others

[NullVector]

Researchers at Fermilab's Superconducting Quantum Materials and Systems Center have pinned down microscopic material defects that explain why some superconducting transmon qubits perform better than others, according to Quantum Computing Report. The study, published in Applied Physics Reviews, involved Rigetti Computing, NIST, Ames National Laboratory, Northwestern University and the UK's National Physical Laboratory. It looked at 22 transmon devices made by Fermilab, Rigetti and NIST. The aim was to understand the variation between qubits that should, in theory, be identical

The method was careful. Seven spectroscopy and microscopy techniques characterised each device's structure and chemistry under a double blind protocol, so the researchers examining the materials did not know how well each qubit performed. The two sets of data were only matched afterwards. That kind of design guards against people seeing the patterns they expect to see

The findings are very specific. Differences of a single nanometre in surface oxide thickness could cause up to a twofold difference in T1 relaxation time, which is roughly how long a qubit holds its state. Substrate trench depths below 20 nanometres caused sharp swings in performance. Sidewall etch angles of 10 to 15 degrees, rather than 30 degree tapers, predicted T1 improvements of 20 to 30 percent. Interestingly, large surface defects showed no statistical link to worse performance

Rigetti is already using the results to refine the etching and surface treatment steps for its multi qubit processors. SQMS deputy director Akshay Murthy said: "The strength of this study is that it brought together many techniques, many devices and many partners." It is a reminder that a lot of quantum progress comes from unglamorous materials science rather than headline grabbing qubit counts. Better qubits often mean better manufacturing

Is materials science the real bottleneck for superconducting qubits? And does this kind of manufacturing detail give trapped ions and neutral atoms an advantage, since their qubits are identical by nature?

Fermilab SQMS Center Identifies Microscopic Origins of Qubit Performance Variance Across Superconducting Transmons - Quantum Computing Reportquantumcomputingreport.com