IonQ tests AI method to prepare problems for quantum computers

IonQ reported progress on a machine-learning method that generates the circuits used to run optimization problems on quantum computers. The work was developed with NVIDIA, Oak Ridge National Laboratory and the University of Tennessee. As the tested problems grew larger, the time needed to generate their circuits stayed roughly constant, according to the report. The approach targets a practical bottleneck: preparing a problem on a conventional computer can be costly enough to undermine the benefit of then using a quantum machine. The reported gain concerns circuit preparation, the step before a quantum calculation runs.