A team from the Julich Supercomputing Centre, AIDAS, RWTH Aachen University, and Purdue University has released a paper that describes a new quantum benchmarking protocol called Linear Ramp Quantum Approximate Optimization Algorithm (LR-QAOA) which they tested on 19 different QPUs from five vendors: IBM, IQM, IonQ,
Quantinuum, and Rigetti for problem sizes requiring from Nq = 5 to Nq = 156 qubits. The tests included three different qubit topologies, 1D-chain, native layout, and fully connected graphs, qubit counts that ranged from 5 to 156, and LR-QAOA layer count (parameter p) that ranged from 3 to 10,000.
It is an interesting study because it provides insights into the strengths and weaknesses of the various processor architecture. The team has posted a technical paper on arXiv showing the results, available here. In addition, Quantinuum has posted a blog discussing the results of their own processors in this study here.
March 22, 2025
Leave A Comment