Extended quantum circuit diagrams
If you are in algorithm development, this work contains a nice summary of interesting rules to optimize quantum circuits. While the ruleset covered is likely incomplete, several interesting tricks our own team enjoys can be found here. Overall, this is a good reference to turn abstract algorithms into real quantum computing programs!
Quantum optimization with globally driven neutral atom arrays
Analog neutral-atom quantum computers are very efficient, but that usually comes at the cost of flexibility for problem encoding. A promising way to mitigate those limitations is to use gadgets, methodically placed extra atoms that encode new logical relations, enriching the space of encodable problems. In previous works, that was done while relying on a degree of site-resolved control over atom energetics. This work develops new methodologies that require only global controls. Furthermore, the long-range interaction between Rydberg atoms may, at times, interfere in the quality of solutions for graph-defined specific problems. The methodologies of this work also help mitigate these limitations, helping advance the field of quantum optimization with neutral-atom analog devices.
Error mitigation and circuit division for early fault-tolerant quantum phase estimation
The age of fault-tolerance arises, but a long road remains ahead. This work develops algorithms for “early fault tolerance”, where trade-offs between partial correction or mitigation of errors via spatial encoding and time overhead may lead to improved results at manageable costs. Importantly, this work extends mitigation techniques to beyond the estimation of mean values of operators and covers relevant algorithms such as quantum phase estimation, relevant to applications in chemistry and other fields.
{{Newsletter-signup}}
Hardness-Dependent Adiabatic Schedules for Analog Quantum Computing
A challenge in analog quantum optimization is the definition of efficient protocols, say, for adiabatic evolution. Usually, variational approaches are used, but these can be costly in time and lead impractical at very large scales. This work considers a new class of protocols that, even without variational optimization, enable problem solution for graphs of considerable hardness, independently of system size. The author shows evidence that the protocol allows for extrapolation and may result in even better performances when combined with hybrid methods at feasible scales.
.svg){kind=small}
