From Quantum Chips to Quantum-AI Data Centers: Building Next-generation Supercomputing with Hybrid Quantum Supercomputers at Scale

Quantum computing is moving out of the laboratory and into the data center — not as a standalone curiosity, but as an accelerator tightly integrated with AI infrastructure. This talk presents Anyon Technologies' full-stack approach to hybrid quantum-AI supercomputing: proprietary modular superconducting quantum processors built on a modular architecture with parametric entangling gates, combined with quantum interconnects, cryogenics, and real-time control electronics, and tightly coupled to NVIDIA GPU systems through NVQLink and CUDA-Q. This architecture recently powered the world's first tightly coupled hybrid quantum-classical commercial data center, deployed with SDT in Seoul. Drawing on lessons from this deployment, the talk examines what it takes to industrialize quantum computing: scaling qubit counts through entangling multiple quantum chips into one big quantum processor, achieving industry-grade development, production, and operation through AI automation, and building out the emerging quantum supply chain — from chip fabrication and advanced packaging to cryogenic components, and control electronics. Taiwan's advanced electronics and semiconductor ecosystem is uniquely positioned to lead in many of these areas, and the talk will map concrete opportunities for collaboration across the quantum-ai-electronics-semiconductor value chain as hybrid quantum-AI systems scale from today's pilots toward million-qubit quantum supercomputers within data centers and, ultimately, networked quantum data centers over a distance.