The biggest thing holding quantum computers back isn't raw power. It's that they can't stay stable long enough to be useful. These machines are incredibly sensitive. Their internal components drift out of alignment constantly, which means researchers run a short calculation, then spend days fixing the machine before they can run another one. It's a brutal cycle.
On April 14, 2026, though, NVIDIA announced something that could break that cycle. At NVIDIA Quantum Day, the company launched NVIDIA Ising, the first open-source family of AI models made specifically for quantum computing. The idea is that instead of researchers manually babysitting quantum hardware, AI does it, catching errors and fixing calibration issues as they happen.
Jensen Huang, NVIDIA's CEO, put it plainly in the announcement: "AI is essential to making quantum computing practical.” He added that, “With Ising, AI becomes the control plane — the operating system of quantum machines.”
Here's what you need to know about the new Ising Open AI models from NVIDIA:
/1. The name comes from physics
Ising is named after the Ising model, a mathematical framework physicists use to simplify how they study complex physical systems. NVIDIA picked the name because this model family does something similar. It takes two of quantum computing's challenging engineering problems and makes them actually manageable.
/2. Ising Calibration reads live data and adjusts in real time
This part of the system watches quantum processors as they run and responds to what it sees, so AI agents can handle the constant tuning that quantum hardware demands. Before this existed, tuning was done by hand and could eat up days of a researcher's time. Ising Calibration shrinks that down to hours and does it without anyone needing to step in.
/3. Ising Decoding comes in two versions
One is built for speed, the other for accuracy. Both handle quantum error correction in real time, and both beat pyMatching, the current open-source standard, by up to 2.5x in speed and 3x in accuracy.
/4. Over 20 institutions are already running it
Ising Calibration is being used at Atom Computing, Academia Sinica, EeroQ, Conductor Quantum, Fermi National Accelerator Laboratory, Harvard's 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 deployed at Cornell University, EdenCode, Infleqtion, IQM Quantum Computers, Quantum Elements, Sandia National Laboratories, SEEQC, UC San Diego, UC Santa Barbara, University of Chicago, University of Southern California, and Yonsei University.
/5. It works with NVIDIA's existing quantum tools
Ising plugs into CUDA-Q, NVIDIA's platform for combining quantum and classical computing, and NVQLink, their hardware interconnect built for real-time quantum control. Developers also get access to NIM microservices and a set of ready-made quantum computing workflows to help them adapt the models to their specific hardware without starting from scratch. Everything runs locally on researchers' own systems, so sensitive data never leaves their infrastructure.
/6. The models are free and available right now
Ising sits alongside NVIDIA's other open models: Nemotron for agentic AI, Cosmos for physical AI, Alpamayo for autonomous vehicles, Isaac GR00T for robotics, and BioNeMo for biomedical research. The models, training data, and frameworks are all up on GitHub, Hugging Face, and build.nvidia.com.
/7. The quantum market is worth $11 billion by 2030
Analyst firm Resonance projects the quantum computing market will cross $11 billion by the end of the decade. That growth depends heavily on progress in error correction and calibration, which happen to be the exact two problems Ising was built to solve.
Ogbonda Chivumnovu
