Nvidia unveils “Ising” quantum AI model
Nvidia has announced a new family of open-source AI models designed to accelerate quantum computing by improving calibration and error correction.
The model, named Ising, is 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.
The company says universities and research institutes have already begun adopting this model for quantum computing development.
Ising uses AI to address the key technical challenges hindering quantum computing, with a focus on improving system reliability rather than relying solely on hardware advances.
Quantum computing has progressed from theory to early practical application, but is still largely in the pre-commercial stage. Companies like Google and IBM, as well as startups like Quantinuum, have demonstrated logical qubits, which are more stable than physical qubits. This is an important milestone on the path to fault-tolerant quantum computers 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, tune qubits, and reduce noise. Many current use cases combine classical AI with quantum computing. While AI handles data-intensive tasks, quantum systems are tested for specific sub-problems such as optimization and simulation.
“AI is essential to making quantum computing practical,” CEO Jensen Huang said in a statement. “At Ising, AI becomes the control plane, or operating system, of quantum machines, transforming fragile qubits into scalable and reliable quantum GPU systems.”
Analyst firm Resonance predicts the quantum computing market will exceed $11 billion in 2030. This growth trajectory is highly dependent on continued progress in addressing key engineering challenges such as quantum error correction and scalability.
What is Ising?
The new Nvidia model is based on a mathematical model from physics that is widely used to represent optimization problems. Basically, Ising models are used to find the optimal solution among many possibilities.
Nvidia introduced Ising to improve how quantum processors tune and manage errors. calibration In this context, it refers to fine-tuning a quantum processor so that the qubits work correctly. error correction It includes detecting and correcting errors that arise from the inherent vulnerabilities of qubits.
The company says its model can perform these tasks faster and more accurately than existing methods.
The goal is to help researchers and companies build quantum systems that can run practical applications.
Nvidia Ising includes customizable models, tools, and data that accelerate quantum processors. These include:
- Ising calibration: A vision language model that can quickly interpret and react to measurements from quantum processors. This allows AI agents to automate continuous calibration, reducing the required time from days to hours.
- Ising decoding: Two variants of a 3D convolutional neural network model. Optimized for either speed or accuracy, it is used to perform real-time decoding for quantum error correction. According to Nvidia, the Ising decoding model is up to 2.5 times faster and three times more accurate than pyMatching, the current open source industry standard.
In true Nvidia fashion, Huang and company aren’t gatekeepers to this technology. By continuing our open model strategy, we will foster ecosystem growth and adopt the same strategy we used to build our AI advantage.
Ising calibrations are already in use at 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 UK National Physical Laboratory.
Ising Decoding has been implemented by Cornell University, EdenCode, Infleqtion, IQM Quantum Computer, Quantum Elements, Sandia National Laboratories, SEEQC, UC San Diego, UC Santa Barbara, University of Chicago, University of Southern California, and Yonsei University.
Why this approach?
Quantum systems are inherently unstable and error-prone, making them difficult to scale. These issues keep 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 control systems can actively manage and stabilize quantum machines, rather than relying solely on hardware improvements.
structure
AI models are used to: Continuously adjust quantum processors to function properly. Detect and fix errors when they occur. Optimize performance across different types of quantum hardware.
This forms part of a hybrid computing approach where traditional computers, AI systems and quantum machines work together to solve problems. Nvidia’s wide range of platforms also rely on GPUs to perform the large-scale computations that support these workloads.
Nvidia has made the model available as an open tool. This means that researchers and companies can use, modify, and build on the model. The company says this could make quantum systems more stable and bring them closer to practical application. The company says the aim is for Ising to show that the future of quantum computing may depend as much on AI software as quantum hardware.
For more information, please visit Nividia’s website.
About the author
John K. Waters is the editor-in-chief of many Converge360.com sites focused on high-end development, AI, and future technologies. He has been writing about Silicon Valley’s cutting-edge technology and culture for more than 20 years and is the author of more than a dozen books. Co-wrote the script for the documentary film Silicon Valley: 100 years of renaissanceaired on PBS. You can contact him at: [email protected].
