Quantum Computing Report

Oxford Ionics Sets New Records for Qubit Gate Fidelities

UK quantum hardware startup, Oxford Ionics, has released performance results from their latest prototype device showing new records for qubit gate fidelity. For single qubit gates they achieved a fidelity of 99.99916% and for two-qubit gates they achieved a fidelity of 99.97%. This is slightly better than the previous record we had seen for two-qubits gates which was 99.914% achieved in April by Quantinuum.

Like a few other companies, the Oxford Ionics devices is based upon a trapped ion technology, but Oxford Ionics uses electronics to control the gate operations of the qubits whereas the others use lasers. This makes it easier to mass produce the qubit chips in a semiconductor foundry and Oxford Ionics already has an an R&D partnership with Infineon Technologies AG to do so. The 10 qubit chip includes seven zones for qubit operations. The company also noted that their design provides low crosstalk between zones and still provided good and stable fidelity performance over a 60 hour testing period. We should also note that their design hasn’t totally eliminated lasers because lasers are used for ion loading, state preparation, and measurement, but not for qubit control.

High levels of qubit fidelity are very important in quantum processors and can potentially be leveraged in both NISQ and FTQC (Fault Tolerant Quantum Computing) applications. For algorithms that only have a moderate depth level of gates, the fidelity may be good enough to provide a usable result without requiring the overhead of error correction. For FTQC systems, better intrinsic fidelity of the physical gate leads to more efficient error correction code implementations which provide more logical qubits for a given number of physical qubits. The company has indicated that they will use the technology implemented in this chip to scale up future designs to 256 qubits.

For more information we refer you to the press release provided by Oxford Ionics here and also a preprint paper posted on arXiv here.

July 12, 2024

Exit mobile version