NVIDIA released NVIDIA Ising Calibration 1.5, an open-source AI model designed to automatically analyze quantum processing unit diagnostic data and determine calibration plans, aiming to automate the quantum computer calibration process. According to ChainCatcher, the company said the visual language model is built for quantum calibration tasks and can interpret quantum chip experimental data, perform zero-shot analysis when no historical cases are available, and use in-context learning with related samples to help quantum devices optimize operating conditions.
In the QCalEval quantum calibration benchmark, Ising Calibration 1.5 reportedly outperformed open-source models of similar size by about 10% in zero-shot reasoning and improved by about 86.5% over the previous generation model when using related experimental cases for in-context learning. It also approached top closed-source large models in performance. The model has 31 billion parameters and supports data center GPUs such as NVIDIA Grace Blackwell and Vera Rubin, while an NVFP4 quantized version can run on a single consumer GPU or NVIDIA DGX Spark.
NVIDIA said the training data covers multiple qubit architectures, including superconducting qubits, quantum dots, ions, neutral atoms, and helium surface electrons. Industry observers said automated calibration remains one of the key bottlenecks to scaling quantum computing, and NVIDIA's tool reflects the expansion of AI from traditional computing into the control layer of quantum hardware.
NVIDIA Releases Open-Source AI Model for Quantum Computer Calibration
2026-07-27 16:23:41
Disclaimer:
1. The information provided does not constitute investment advice. Investors should make independent decisions and bear all risks themselves.
2. The copyright of this content belongs to the original author. The views expressed herein are solely those of the author and do not represent the stance or position of this website.
