NVIDIA announced in April 2026 the world's first family of open-source quantum AI models, NVIDIA Ising, designed to accelerate the development of quantum processors capable of running useful applications. The models deliver quantum error-correction decoding that is up to 2.5x faster and 3x more accurate than traditional approaches, addressing two of the most significant bottlenecks in quantum computing.
Two-Part Model Family Targets Critical Quantum Challenges
The NVIDIA Ising family consists of two main components addressing distinct quantum computing challenges:
Ising Calibration: A 35 billion parameter vision-language model trained on multi-modality qubit data. This component automates the rapid tuning of quantum processors, supporting agentic calibration automation. Quantum processors require constant recalibration as qubits are highly sensitive to environmental factors, and manual calibration is time-consuming and error-prone.
Ising Decoding: Accelerates the real-time decoding required for quantum error correction. Quantum computers require continuous error correction because qubits are fragile and prone to errors from environmental interference. Traditional classical algorithms for error correction are computationally intensive and slow, limiting the practical use of quantum computers.
AI Replaces Classical Algorithms for Quantum Optimization
Quantum computing has been constrained by the computational difficulty of calibrating quantum processors and correcting errors in real-time. NVIDIA's approach replaces classical algorithms with AI models that dramatically improve both speed and accuracy. According to Tom's Hardware, the Ising models are 2.5x faster and 3x more accurate than existing tools for error-correction decoding.
The models are fully open-source on GitHub, enabling researchers and quantum computing companies to integrate them into their workflows without vendor lock-in. This represents a significant shift from proprietary quantum optimization tools that have dominated the industry.
Major Research Institutions and Quantum Companies Adopt Ising
Leading quantum enterprises, academic institutions, and research laboratories adopting NVIDIA Ising include:
- Academia Sinica
- Fermi National Accelerator Laboratory
- Harvard John A. Paulson School of Engineering and Applied Sciences
- Infleqtion
- IQM Quantum Computers
- Lawrence Berkeley National Laboratory's Advanced Quantum Testbed
- U.K. National Physical Laboratory (NPL)
This release is part of NVIDIA's broader strategy to provide open-source AI tooling across multiple domains. NVIDIA has positioned itself as a key enabler of the quantum computing industry, providing both hardware (GPUs for quantum simulation) and now AI software for quantum processor optimization.
Key Takeaways
- NVIDIA Ising is the world's first family of open-source AI models designed specifically for quantum computing, offering 2.5x faster and 3x more accurate error-correction decoding than traditional approaches
- The Ising Calibration model uses 35 billion parameters trained on multi-modality qubit data to automate quantum processor calibration
- The models are fully open-source, allowing quantum computing companies and researchers to integrate them without vendor lock-in
- Major adopters include Fermi National Accelerator Laboratory, Harvard, Lawrence Berkeley National Laboratory, IQM Quantum Computers, and the U.K. National Physical Laboratory
- NVIDIA's approach uses AI to replace classical algorithms for quantum processor calibration and error correction, addressing two critical bottlenecks in quantum computing development