NVIDIA shared in a release that the organization is launching NVIDIA Ising, the first family of open-source quantum AI models designed to accelerate the development of reliable quantum processors. By providing open-access tools, the firm allows researchers to maintain control over their data while utilizing high-performance AI to stabilize fragile qubits.
Jensen Huang, Founder and CEO of NVIDIA, stated, “With Ising, AI becomes the control plane—the operating system of quantum machines—transforming fragile qubits to scalable and reliable quantum-GPU systems.” These models integrate with existing platforms like CUDA-Q to create a comprehensive environment for hybrid quantum-classical computing.
The Ising family includes specialized variants for calibration and decoding that significantly improve the speed and accuracy of quantum operations. The calibration model utilizes vision language technology to automate processor measurements, reducing task durations from days to hours. Meanwhile, the decoding models employ 3D convolutional neural networks to perform real-time error correction with up to 3x higher accuracy than current industry standards. Leading academic institutions and research labs, such as Harvard and Sandia National Laboratories, already use these tools to tackle complex problems in quantum physics.