Upgrading AI Computing Infrastructure into Accelerated Quantum-AI Supercomputers with NVIDIA NVQLink and CUDA-Q

Open, low-latency orchestration between quantum systems and accelerated computing advances logical-qubit scale and utility-scale quantum applications.

Singapore & Berkeley, Calif., October 28, 2025— Anyon Technologies (“Anyon”), a vertically integrated quantum computer and hybrid quantum–classical computing system company, announced support for NVIDIA NVQLink with NVIDIA CUDA-Q to enable high-performance computing (HPC) and AI data centers to extend their existing AI infrastructure into future accelerated quantum-AI supercomputers. Anyon’s quantum computers integrate with NVIDIA accelerated computing via NVQLink’s open system architecture to accelerate calibration, decoding, networking, and other classical-compute-intensive control tasks on GPUs that are tightly coupled to Anyon’s QPUs in real time.

“Our view is simple: in the future, all supercomputers will include quantum computers with ultra-low error logical qubits—and building those logical qubits at scale requires tight GPU-QPU integration via a technology like NVQLink,” added Dr. Roger Luo, co-founder & CEO, Anyon Technologies. “This collaboration lets Anyon’s customers and end-users get ready for the impacts to come and pursue real-world applications faster while preserving architectural flexibility as the physical interface between GPU and QPU evolves under NVQLink standards.”

A Quantum-leap for Computing Infrastructures

• Leverage existing AI investments. Deploy Anyon QPUs and compatible control stacks alongside current CPU/GPU nodes and RDMA over Converged Ethernet (RoCE) fabrics; run hybrid jobs through familiar DC workflows.

• Future-proof, open interface. NVQLink and CUDA-Q work together to provide an open, interoperable workflow that evolves with GPU generations, compatible control electronics (FPGA/ASIC/cryo-electronics), and Anyon’s QPU iterations.

• Tightly integrated hybrid performance. Real-time, low-latency high-throughput orchestration between QPUs and accelerated computing enables classical-compute-intensive pipelines, such as machine learning routines, error decoding and dynamic control, needed for scaling quantum computing with ultra-low error logical qubits using quantum error correction and modular quantum computing at scale.

• Scale from single node to multi-node. Starting with a single “real-time host”, Anyon will be able to expand to multi-node GPU clusters as systems and demand grow. The tight coupling between advanced GPU/CPU supercomputing and QPU nodes further enables coherent quantum computing to expand from a single quantum computer node to a cluster of quantum nodes within the computing infrastructure as more Anyon's quantum computers are linked with Anyon's quantum interconnect technology.

“With NVQLink and CUDA-Q, customers can add quantum computers to the GPU servers and RoCE networks they already run—protecting today’s investments while building toward accelerated quantum supercomputers,” said Dr. Luo. “This open, deterministic hybrid interface lets operators start small and scale out using the same patterns they trusted for scaling up GPU clusters for AI.”

Enable Quantum Applications for Enterprise End-users

End-users can immediately access tightly coupled, low-latency, and scalable hybrid quantum–classical computing resources from Anyon through their HPC/AI computing resource providers to drive enterprise application development and even deployment now—from finance to chemistry, logistics, and AI-assisted workflows. Tightly coupled classical–quantum computing can unlock breakthroughs in AI and other high-impact applications via algorithms that interleave QPU operations with GPU-accelerated learning/inference in real time.

“What matters is impact today,” Dr. Luo added. “Anyon working with NVQLink lets enterprises build, test, and deploy tightly coupled quantum-classical applications on the same data-center platforms they already trust—opening the door to AI breakthroughs and other mission-critical wins.”

Create Accelerated Quantum-AI Supercomputers

Anyon’s path to a large-scale, error-corrected quantum supercomputer centers on quantum interconnects between Anyon's QPUs and ultra-low-latency interactions between QPU control and GPU/CPU. Anyon modular quantum computing technology targets entangling hundreds of thousands of superconducting qubits distributed across hundreds of quantum computers over a local area entanglement network (qLAN) inside an HPC or AI data center.

• Active measurements + dynamic feed-forward create a robust entanglement network between quantum processor modules, orchestrated through NVQLink-defined, low-latency deterministic links to accelerated classical compute on GPU/CPU systems.

• This design leverages the powerful AI supercomputing fabric and programmability to reach the flexibility, scale, and reliability demanded as the number of entangled quantum modules and ultra-low-error logical qubits grows towards 1,000.

• By combining heavy dynamic control, decoding, and machine learning on GPUs with Anyon’s advances in QPUs and quantum interconnects, we deliver a practical, incrementally scalable route from today’s pilots to tomorrow’s accelerated quantum supercomputers.

“NVIDIA NVQLink provides the tight, deterministic coupling between the quantum stack and accelerated classical compute needed to scale quantum hardware and run hybrid applications”, said Tim Costa, General Manager for Quantum at NVIDIA, “By leveraging NVQLink, Anyon is building a single open interface for practical quantum applications and hybrid quantum-classical workflows”

About Anyon Technologies

Anyon Technologies builds superconducting quantum systems with proprietary quantum interconnects and quantum processors (QPUs) integrated with accelerated computing to develop hybrid quantum–AI computing systems and enable the next-generation supercomputing platforms for enterprise and research. Operating across Singapore, South Korea, and the U.S., Anyon partners with leading labs, companies, and data-center providers on delivering utility-scale accelerated quantum-AI supercomputing resources to the world and enabling end-user driven development of transformative quantum applications with real-world impacts.