Quantum Circuits Links NVIDIA CUDA‑Q with Dual‑Rail Qubits, Enabling Apps

Rob Schoelkopf from Yale University has highlighted the significance of Quantum Circuits’ integration of NVIDIA CUDA‑Q into its Aqumen software suite, marking the first dual‑rail qubit application to incorporate built‑in error detection. The partnership enables users to prototype and test CUDA‑Q applications on AquSim, a simulator that precedes deployment on the company’s full‑stack quantum computing systems, including the Aqumen Seeker QPU – the world’s first system to offer error detection that dramatically improves quantum calculation with industry‑leading fidelity and superconducting execution speed. By combining accelerated computing with dual‑rail superconducting qubits, the collaboration promises to accelerate the path to commercial‑ready quantum computing, a goal echoed by Tim Costa, senior director of Quantum and CUDA‑X at NVIDIA, and Ray Smets, CEO of Quantum Circuits.

Quantum Circuits Integrates CUDA Q With Dual Rail Qubits To Enable Hybrid Workflows

On 2 September 2025, Quantum Circuits, Inc., based in New Haven, CT, announced the integration of NVIDIA’s CUDA‑Q programming framework with its Aqumen software suite. CEO Ray Smets and NVIDIA senior director Tim Costa of Quantum and CUDA‑X led the partnership, which enables developers to write and run hybrid quantum‑classical applications on the company’s proprietary dual‑rail qubits. These qubits encode logical information across two superconducting circuits, providing on‑chip error detection without ancillary qubits, thereby reducing physical qubit overhead and boosting logical fidelity. The Aqumen Seeker quantum processing unit, the first commercial system to employ this architecture, is supported by the AquSim high‑performance simulator and offers a complete set of universal quantum gates, real‑time feedforward, and conditional operations essential for hybrid workflows. Quantum Circuits, founded by quantum physicists including Yale professor Rob Schoelkopf, has opened a strategic quantum release programme inviting enterprises and high‑performance computing developers to access CUDA‑Q applications on its hardware, while the dual‑rail design allows the processor to sustain longer coherent sequences.

The Aqumen Seeker quantum processing unit, unveiled alongside the announcement, is the first commercial system to employ dual‑rail qubits with on‑chip error detection, delivering logical fidelity that surpasses conventional single‑qubit designs while maintaining superconducting execution speeds that meet industry benchmarks. Aqumen Seeker achieves cycle times that match or surpass the fastest superconducting platforms currently available, allowing it to keep pace with high‑performance computing workloads. The combination of rapid gate operations and robust error detection enables the device to handle complex quantum algorithms.

Tim  Costa, senior director of Quantum and CUDA‑X at NVIDIA, underscored that the partnership accelerates hybrid quantum‑classical applications for enterprise high‑performance computing developers by making dual‑rail qubits directly accessible to CUDA‑Q code. He described the collaboration as the first practical platform for researchers to harness dual‑rail superconducting qubits within a CUDA‑Q environment and outlined a practical pathway: write CUDA‑Q applications in the Aqumen suite, validate them in AquSim, and deploy them on the Aqumen Seeker. The initiative is expected to shorten the timeline to fault tolerance and enable enterprise developers to integrate quantum acceleration into existing HPC workflows with unprecedented speed and reliability.