Quantum hits the boardroom
NVIDIA’s NVQLink and related GTC panels signaled quantum computing moving from lab curiosity to board-level strategy — boards must now consider IP risk, talent, and audit committee technical literacy as quantum becomes operationally relevant. The conversation at GTC framed quantum as an inflection point that will change both opportunity and oversight needs. (youtube.com)
NVIDIA formally introduced NVQLink as an open system architecture at GTC on Oct. 28, 2025, positioning the link as a low‑latency bridge between GPUs and quantum processors to build hybrid supercomputers. (marketscreener.com) NVIDIA’s launch materials and follow‑on releases list participation from 17 QPU builders, five controller builders and nine U.S. national labs, naming Brookhaven, Fermi, Lawrence Berkeley, Los Alamos, MIT Lincoln Laboratory, Oak Ridge, PNNL and Sandia among collaborators. (marketscreener.com) Multiple vendors and labs announced NVQLink integrations at and around GTC: IQM and Zurich Instruments unveiled a real‑time QEC demonstrator, Atom Computing reported NVQLink integration for its control stack, Infleqtion showcased its Sqale QPU on the NVQLink booth, and Quantum Machines released an NVQLink integration for its orchestration platform. (marketwatch.com) Technical briefs released with NVQLink cite microsecond‑scale transfers using RDMA‑over‑Ethernet for deterministic, low‑latency movement, integration with NVIDIA’s CUDA‑Q real‑time APIs, and demonstrations pairing NVQLink with GH200/Grace Hopper superchip servers for control and calibration workloads. (infleqtion.com) At GTC demos, a UCL‑led pipeline ran a 54‑qubit IQM system alongside 120 NVIDIA H100 GPUs for hybrid biomolecular simulation, while IQM’s partner releases and Atom Computing materials described NVQLink deployments aimed at accelerating quantum error correction and scaling control systems. (ucl.ac.uk) NVIDIA and partners framed NVQLink as upgradeable at the system level—Quantum Machines said existing DGX Quantum customers can adopt NVQLink without hardware changes—and NVIDIA executives stated “every NVIDIA GPU scientific supercomputer will be hybrid,” underscoring the company’s strategic roadmap presented during GTC’s Quantum Day sessions on March 20, 2026. (quantum-machines.co)
