Qblox Enables Real-Time Quantum Acceleration with NVIDIA CUDA-Q Integration

Qblox is enabling faster quantum processing by directly linking its control systems with NVIDIA’s accelerated computing platform through the cudaq-realtime API. This integration establishes a high-speed connection between quantum computers and graphics processing units, minimizing latency to a few microseconds and addressing a key bottleneck in Variational Quantum Algorithms. The architecture allows Oxford Quantum Circuits’ quantum processors to work more closely with GPU acceleration, potentially advancing research, calibration, and real-world applications in fields like finance and materials science. “Realizing the full potential of quantum computing requires tight integration with classical accelerated computing,” said Simon Phillips, CTO of Oxford Quantum Circuits. According to Qblox CEO Niels Bultink, this work “lays the foundation for scalable, utility-driven applications,” marking a step toward practical quantum-accelerated workflows. NVQLink and CUDA-Q Enable Microsecond Quantum-GPU Feedback A communication bottleneck previously hindering progress in quantum computing is now being addressed through the integration of Qblox control systems with NVIDIA’s NVQLink and CUDA-Q platform. This pairing achieves feedback loops measured in just a few microseconds. This direct, high-speed connection between NVIDIA accelerated computing and full-stack quantum systems, specifically those from Oxford Quantum Circuits (OQC), promises to significantly improve the performance of Variational Quantum Algorithms (VQA) which depend on rapid data exchange. Qblox engineered its control electronics to minimize latency by leveraging NVIDIA’s cudaq-realtime API, a new real-time FPGA-to-GPU communication tool, allowing OQC’s quantum processors to synchronize more effectively with GPU acceleration for research, calibration, and potential utility-scale applications. Qblox Control Stacks Accelerate Oxford Quantum Circuits Processors The pursuit of practical quantum computation increasingly focuses on bridging the gap between quantum processors and classical high-performance computing; current systems often suffer from communication bottlenecks that limit the speed and efficiency of complex algorithms. Qblox, a provider of quantum control stacks, has addressed this challenge through integration with NVIDIA’s NVQLink and the cudaq-realtime API, establishing a direct, high-speed connection to Oxford Quantum Circuits’ (OQC) full-stack quantum computers. This architecture minimizes latency, enabling feedback loops within a few microseconds, a critical improvement for Variational Quantum Algorithms (VQA) where rapid data exchange is essential. The collaboration, building on work initiated in October 2025, aims to synchronize quantum processing with GPU acceleration for research, calibration, quantum error correction, and eventual utility-scale applications spanning finance, chemistry, and materials science. Qblox’s control electronics are specifically designed to leverage NVIDIA’s real-time FPGA-to-GPU communication, a feature enabled by the cudaq-realtime API, to achieve this low-latency performance. Qblox will demonstrate this real-time GPU-to-quantum integration at the NVIDIA GTC conference, booth number 444, offering attendees a chance to discuss hybrid quantum-classical infrastructure. Realizing the full potential of quantum computing requires tight integration with classical accelerated computing. Simon Phillips, CTO of Oxford Quantum Circuits (OQC) Source: https://www.prnewswire.com/news-releases/qblox-powers-real-time-quantum-acceleration-with-nvidia-cuda-q-integration-302715116.html CUDA-Q NVIDIA OXFORD QUANTUM CIRCUITS QBLOX QUANTUM COMPUTING