ORCA Computing has collaborated with the Poznań Supercomputing and Networking Center (PCSS) and Imperial College London to demonstrate the integration of photonic quantum processors into an operational data center environment. The collaboration, which utilized NVIDIA accelerated computing and the NVIDIA CUDA-Q software platform, established a working distributed quantum neural network. This effort provides a reference architecture for deploying scalable hybrid quantum–classical systems for AI and machine learning workloads.
The demonstration was executed within the PCSS data center, where two ORCA Computing PT-1 photonic quantum processors were deployed alongside existing NVIDIA Hopper GPUs. Operating at room temperature, the processors were integrated into the classical high-performance computing (HPC) environment through the CUDA-Q platform and the Slurm workload manager. This unified software stack enabled researchers to program both classical and quantum resources within a single environment. This marks the first time photonic quantum processors have been integrated through CUDA-Q.
This demonstration highlights the robustness of photonic quantum processors under distributed operation and their ability to support workloads that are naturally aligned with modern HPC environments. The collaboration leveraged research from Imperial College, which was honored as “Best Paper” in the photonic track at IEEE Quantum Week. By running a distributed system in a data center, the collaboration provides a proof of viability and a replicable framework for integrating quantum enhancements into existing HPC environments.
Read more about this collaboration in the ORCA Computing press release here, the ORCA Computing blog post here, the NVIDIA press release here, and our related QCR coverage of the OQC, Digital Realty, and NVIDIA Quantum-AI Data Centre launch here.
September 17, 2025
