NVIDIA Ising quantum computing models have been released as the first open-source family of artificial intelligence models designed to accelerate the development of practical quantum processors. The announcement came on April 15, 2026, from Santa Clara, California.
The NVIDIA Ising quantum framework addresses two critical challenges in quantum computing: calibration of quantum processors and error correction. Researchers and companies can now access artificial intelligence tools to build scalable, high-performance quantum systems while maintaining full control over their data and infrastructure.
Performance Improvements in NVIDIA Ising Quantum Systems
The Ising family delivers measurable performance gains over existing methods. Decoding models achieve 2.5 times faster speed and three times higher accuracy compared to pyMatching, the open-source industry standard currently in use. The calibration model is 15 times smaller than competing alternatives while maintaining real-time responsiveness to quantum processor measurements.
Jensen Huang, NVIDIA founder and chief executive, stated that artificial intelligence serves as the operating system for quantum devices. The technology transforms fragile quantum bits into reliable, scalable quantum processing units capable of running practical applications.
“Artificial intelligence is essential to making quantum computing practical. With Ising, AI becomes the control system for quantum hardware, enabling the transformation of fragile qubits into more reliable and scalable quantum processing units.”
Jensen Huang, Founder and Chief Executive Officer, NVIDIA
NVIDIA Ising Quantum Model Components
The Ising program includes two specialized models. Ising for Calibration uses vision and language capabilities to interpret measurements from quantum processors and enable continuous automation of calibration processes. Ising for Decoding comprises two three-dimensional convolutional neural network variants optimized for real-time decoding during quantum error correction operations.
NVIDIA provides comprehensive workflow documentation, training data, and NIM microservices that allow developers to fine-tune models for different hardware architectures and use cases. The models run locally on researcher systems, ensuring protection of sensitive data.
Adoption by Leading Research Institutions
Major academic institutions and research laboratories have already adopted the NVIDIA Ising quantum framework. Organizations using Ising for calibration include Atom Computing, Academia Sinica, EeroQ, Conductor Quantum, Fermi National Accelerator Laboratory, Harvard John A. Paulson School of Engineering and Applied Sciences, Infleqtion, IonQ, IQM Quantum Computers, Lawrence Berkeley National Laboratory’s Advanced Quantum Testbed, Q-CTRL, and UK National Physical Laboratory.
Institutions deploying Ising for decoding include Cornell University, EdenCode, Infleqtion, IQM Quantum Computers, Quantum Elements, Sandia National Laboratory, SEEQC, University of California San Diego, UC Santa Barbara, University of Chicago, University of Southern California, and Yonsei University.
Integration with NVIDIA Quantum Computing Platform
The NVIDIA Ising quantum system integrates with the CUDA-Q programming platform for hybrid quantum-classical computing and NVQLink technology for device connectivity between CPUs and GPUs. This integration enables real-time quantum error correction and control, providing researchers with a complete toolkit to transition current quantum bits into future high-speed quantum computers.
The quantum computing market is projected to exceed 11 billion dollars by 2030, according to research firm Resonance, driven by continued progress in addressing fundamental engineering challenges such as quantum error correction and scalability. Open-source models like Ising accelerate this development by enabling broader access to artificial intelligence capabilities across the research community.
The NVIDIA Ising quantum models are available alongside training data and frameworks through GitHub, Hugging Face, and build.nvidia.com. The release joins NVIDIA’s portfolio of open models including Nemotron for agent systems, Cosmos for physics-based AI, Alpamayo for autonomous vehicles, Isaac GR00T for robotics, and BioNeMo for biomedical research.
