News

Nvidia Announces 'Ising' Quantum AI Model

• By John K. Waters
• 04/20/2026

Nvidia announced a new family of open-source AI models designed to accelerate quantum computing by improving calibration and error correction.

Dubbed Ising, the models are designed to deliver up to 2.5x faster and 3x more accurate quantum error correction decoding while enabling automated calibration workflows that reduce setup time from days to hours, the company said.

Universities and research labs have already begun adopting the models for quantum computing development, the company said.  

Ising uses AI to address the main technical challenges holding back quantum computing, focusing on improving system reliability rather than relying solely on hardware advances. 

Quantum computing is moving from theory toward early practical use, but it's still largely in a pre-commercial phase. Companies like Google and IBM, as well as startups such as Quantinuum, have demonstrated logical qubits that are more stable than physical ones. This is a key milestone on the road to fault-tolerant quantum computers, which are needed for useful, large-scale applications.

AI and quantum computing are starting to reinforce each other. Machine learning is being used to design better quantum hardware, calibrate qubits, and reduce noise. Many current use cases combine classical AI with quantum computing. AI handles data-intensive tasks, while quantum systems are tested on specific subproblems such as optimization or simulation.

"AI is essential to making quantum computing practical," said CEO Jensen Huang, in a statement. "With Ising, AI becomes the control plane—the operating system of quantum machines—transforming fragile qubits to scalable and reliable quantum-GPU systems." 

Analyst firm Resonance expects the quantum computing market to surpass $11 billion in 2030. This growth trajectory is highly dependent on continued progress in addressing critical engineering challenges, such as quantum error correction and scalability. 

What is Ising?
The new Nvidia models are based on a mathematical model from physics that is widely used to represent optimization problems. Essentially, Ising models are used to find the best solution among many possibilities. 

Nvidia introduced Ising to improve how quantum processors are calibrated and errors are managed. Calibration in this context refers to fine-tuning a quantum processor so that its qubits behave correctly, while error correction involves detecting and correcting errors that arise from qubits' inherent fragility.

The company says the models can perform these tasks faster and more accurately than existing methods. 

The aim is to help researchers and companies build quantum systems capable of running practical applications.

Nvidia Ising includes customizable models, tools and data that accelerate quantum processors. These include:  

• Ising Calibration: A vision language model that can speedily interpret and react to measurements from quantum processors. This enables AI agents to automate continuous calibration, reducing the time needed from days to hours. 
• Ising Decoding: Two variants of a 3D convolutional neural network model optimized for either speed or accuracy and used to perform real-time decoding for quantum error correction. Ising Decoding models are up to 2.5x faster and 3x more accurate than pyMatching, the current open-source industry standard, according to Nvidia. 

In true Nvidia fashion, Huang and company are not gatekeeping this technology. By continuing with the open model strategy, it encourages ecosystem growth and adopts the same playbook it used to build AI dominance.  

Ising Calibration is already in use by 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 the U.K. National Physical Laboratory. 

Ising Decoding is being deployed by Cornell University, EdenCode, Infleqtion, IQM Quantum Computers, Quantum Elements, Sandia National Laboratories, SEEQC, University of California San Diego, UC Santa Barbara, University of Chicago, University of Southern California , and Yonsei University. 

Why this approach?
Quantum systems are difficult to scale because they are inherently unstable and error-prone. These issues have kept most quantum computers in the experimental stage. 

Nvidia's approach is based on the idea that machine learning systems trained to predict errors, optimize performance, and control systems can actively manage and stabilize quantum machines, rather than relying solely on hardware improvements. 

How it works
The AI models are used to: continuously adjust quantum processors so they function correctly; detect and correct errors as they occur; and optimize performance across distinct types of quantum hardware.  

This forms part of a hybrid computing approach in which traditional computers, AI systems, and quantum machines work together to solve problems. Nvidia's broader platform also relies on GPUs to perform large-scale calculations that support these workloads. 

Nvidia has made the models available as open tools, meaning researchers and companies can use, modify, and build on them. The company says this could help make quantum systems more stable and closer to practical use. The company says its aim is for Ising to show that the future of quantum computing may depend as much on AI software as on quantum hardware.