New AI models address quantum computing challenges, improving calibration and error correction while enabling faster processing and accurate systems for real use.
NVIDIA has introduced a new family of open-source quantum AI models called Ising to improve how quantum processors are calibrated and corrected for errors. The models address two key challenges in quantum computing and can deliver up to 2.5 times faster processing and up to 3 times better accuracy in error decoding.
Quantum systems require continuous calibration and error correction to run reliably at scale. These processes are complex and time-consuming. AI is now being used to automate both, helping improve system performance and reduce manual effort. Open-source access also allows developers to control their data and infrastructure while building and testing quantum systems.
The Ising models are based on a mathematical framework used to study complex physical systems. NVIDIA applies this concept to develop AI tools focused on calibration and quantum error correction, which are important for hybrid systems that combine classical and quantum computing.
The offering includes models, tools, and data designed for quantum workflows. Ising Calibration is a vision-language model that reads measurement data and responds in real time, reducing calibration time from days to hours. Ising Decoding includes two versions of a 3D convolutional neural network for error correction, one focused on speed and the other on accuracy, both improving performance compared to existing methods like pyMatching.
“AI is essential to making quantum computing practical,” said Jensen Huang, founder and CEO of NVIDIA. “With Ising, AI becomes the control plane — the operating system of quantum machines — transforming fragile qubits to scalable and reliable quantum-GPU systems.”
The models are being used by research and industry groups, including Atom Computing, IonQ, IQM Quantum Computers, and Fermi National Accelerator Laboratory, to build and improve quantum processors and hybrid computing systems.
Click here for the original announcement.
