NVIDIA Ising Cuts Quantum Calibration From Days to Hours - And It's Open Source

NVIDIA has released Ising, an open-source family of AI models designed to solve quantum computing's most pressing bottlenecks. The initiative targets two critical challenges: quantum processor calibration and real-time error correction, both essential for making quantum computers practically viable. The Ising suite includes a vision language model that automates quantum processor calibration, reducing what previously required days of manual adjustment to just hours of automated processing. Additionally, it features a 3D convolutional neural network for real-time quantum error correction decoding, offered in two variants optimised for either speed or accuracy depending on specific use case requirements. Performance metrics demonstrate significant improvements over existing solutions, with error correction decoding that operates up to 2.5 times faster whilst achieving three times greater accuracy than current methods. The open-source nature ensures immediate global accessibility for quantum research teams. Major research institutions have already adopted the technology, including Harvard University, Fermilab, Lawrence Berkeley National Laboratory, IQM, Infleqtion, and the UK's National Physical Laboratory. This widespread early adoption signals NVIDIA's strategic positioning beyond hardware provision into software infrastructure that bridges classical and quantum computing paradigms, potentially establishing dominance in the emerging quantum ecosystem toolchain.