ORCA Computing Accelerates Photonic Simulation with NVIDIA cuTensorNet

ORCA Computing has integrated the NVIDIA cuTensorNet library and the cuQuantum SDK to accelerate photonic quantum simulation and support scalable hybrid workflows. This technical integration addresses a gap in the simulation ecosystem, which has historically prioritized qubit-based models over the unique requirements of photonic circuit architectures. By utilizing GPU-accelerated tensor network methods, ORCA can now model larger and more complex photonic circuits with higher computational efficiency than previously possible on CPU-only systems.

These simulation capabilities are specifically optimized for ORCA’s PT-2 processor, a rack-mounted, room-temperature photonic quantum computer designed for standard data center environments. The simulator allows developers to prototype algorithms and validate architectures in a virtual environment that mirrors the performance of the PT-2 hardware. This GPU-driven approach facilitates the benchmarking of hybrid quantum-classical systems, enabling researchers to tightly couple photonic quantum routines with classical high-performance computing (HPC) tasks within the NVIDIA CUDA ecosystem.

ORCA plans to open-source this photonic simulator in alignment with a future release of NVIDIA CUDA-Q to encourage community-driven benchmarking and development. This move follows the successful deployment of ORCA systems at the UK National Quantum Computing Centre (NQCC) and the Poznan Supercomputing and Networking Center (PSNC). By providing open access to hardware-aligned simulation tools, the collaboration aims to lower the barrier for organizations integrating photonic quantum processors into existing enterprise AI and optimization stacks.

For full technical details on the cuTensorNet integration and the open-source simulator roadmap, consult the official HPCwire announcement here.

March 16, 2026