NVIDIA integrates quantum computing via NVQLink and IonQ deployment

NVIDIA is expanding its role in the quantum computing sector through the introduction of NVQLink, an open architecture interconnect designed to bridge quantum processing units (QPUs) with GPU-accelerated supercomputers. The technology aims to facilitate hybrid AI and quantum workloads by providing high throughput and low latency between hardware components.

As part of this ecosystem, IonQ has announced the deployment of its Superion 256 quantum computer at NVIDIA’s Accelerated Quantum Research Center. This on-premise system utilizes NVQLink and CUDA-Q to connect IonQ hardware to NVIDIA’s GB200 NVL72 infrastructure.

NVIDIA has assembled a coalition of 17 QPU builders and five controller vendors around the NVQLink standard. Early performance metrics show that Quantinuum’s Helios QPU, when connected via NVQLink to an NVIDIA GH200 Grace Hopper Superchip, achieved real-time quantum error correction decoding significantly faster than standard requirements. Adoption is growing, with over a dozen supercomputing centers globally, including nine US national laboratories, already working with the technology.