In this webinar, researchers from QuEra and NVIDIA present their recent work on QC-Daemon, an AI-based agent designed for move synthesis in reconfigurable neutral atom quantum processors.

The session explores QuEra’s newly introduced zoned architecture, which leverages reconfigurable atom arrays, and shows how the constraints of atom movement can be formulated as a graph optimization problem. Using reinforcement learning, the team developed an agent that efficiently maps abstract quantum circuits onto hardware, while dynamically optimizing atom layouts for the future cost of reconfiguration and gate execution.