IQM Automates Quantum Calibration with NVIDIA Ising to Support Enterprise Adoption

IQM Quantum Computers has announced the implementation of AI-driven agentic calibration, a system designed to automate the tuning of superconducting quantum processors using the NVIDIA Ising open model family. Revealed on World Quantum Day, this development aims to eliminate the "manual bottleneck" of sequential calibration by using visual AI agents to inspect and tune multiple qubits simultaneously. This parallelized approach is intended to make quantum systems operationally viable for high-performance computing (HPC) data centers and "AI factories" by reducing the requirement for on-site quantum engineering specialists.

The core of the architecture utilizes NVIDIA Ising Calibration models, which are fine-tuned to interpret measurement plots and adjust hardware parameters in real-time. This automation shifts the maintenance burden from human experts to intelligent agents, addressing the global scarcity of quantum talent that has previously hindered the deployment of on-premises quantum infrastructure. According to Juha Vartiainen, Co-founder of IQM, the goal is to allow enterprises to focus on application-level results rather than the underlying infrastructure maintenance.

This collaboration builds on the existing integration between IQM and the NVIDIA quantum stack, specifically the NVIDIA NVQLink hardware interconnect and the CUDA-Q hybrid software platform. By integrating AI agents directly into its calibration infrastructure, IQM aims to maintain higher algorithmic efficiency and consistent high-fidelity performance across its superconducting QPUs. This production-focused strategy is designed to provide institutions with the capability to own and operate quantum systems at scale, bridging the gap between experimental laboratory setups and industrial-grade computing environments.

For the official press release regarding IQM’s agentic calibration and its integration with NVIDIA Ising, visit the IQM newsroom here. Technical details on the NVIDIA Ising open model family and its role in quantum-GPU supercomputing can be found via the NVIDIA newsroom here.

April 15, 2026