Quantum Machines runs NVIDIA CUDA-Q across live qubits and GPUs

Quantum Machines said it became the first control company to execute an end-to-end NVIDIA CUDA-Q program across live qubits and a classical processor using NVQLink. The demonstration, shown at IEEE Quantum Week in Toronto, is aimed at making hybrid quantum-classical development faster and easier for software teams.

Why it matters: - The demonstration points to a simpler way for developers to build hybrid quantum-classical applications without hand-coding low-level control sequences. - Faster coordination between qubits, GPUs and CPUs could shorten time-to-value for research and commercial quantum workloads. - The microsecond-speed link is especially important for future uses such as quantum error correction.

What happened: - Quantum Machines became the first control company to run an end-to-end NVIDIA CUDA-Q program across live qubits and a PPU classical processor with NVIDIA NVQLink. - The company is showing the live run this week at IEEE Quantum Week in Toronto. - The demo used code written once in CUDA-Q and executed it across a quantum processor, GPUs and CPUs as one system.

The details: - CUDA-Q is NVIDIA’s open platform for quantum-GPU supercomputing. - CUDA-Q lets developers write instructions that both classical computers and quantum computers can follow. - The workflow can be written in familiar languages used across the quantum industry, including Python, C++ and QUA. - NVQLink is a microsecond-speed connection architecture between a quantum controller and GPUs. - Quantum Machines integrated NVQLink into its Orchestration Platform. - The integration connects the hardware that drives and reads qubits directly to NVIDIA accelerated computing through a low-latency link to classical processors. - Quantum Machines’ control system converts CUDA-Q operations into precisely timed signals that control and measure the qubits. - The full exchange completed in about a millionth of a second. - The system routes each part of the workload to the appropriate processor.

Between the lines: - The demonstration is meant to reduce the specialized hardware expertise traditionally needed to deploy quantum workloads. - The setup moves quantum computing closer to a model where QPUs act like another compute resource applications can call when needed. - NVIDIA framed the work as part of a future unified quantum supercomputing system. - Quantum Machines and NVIDIA have a long history of low-latency QPU-GPU integration, and NVQLink is the latest step in that effort.

What's next: - Quantum Machines is positioning the stack for scalable, application-driven quantum deployments. - The company and NVIDIA are pushing toward real-time quantum-classical interactions required for larger workloads. - Future development is likely to focus on deeper integration between qubits, GPUs and CPUs for practical quantum computing applications.

The bottom line: - Quantum Machines is trying to make quantum programming look more like mainstream software development, with NVIDIA tools handling much of the hard wiring between classical and quantum systems.