NVIDIA Ising 1.5: Quantum Calibration with In-Context Learning

NVIDIA has released Ising Calibration 1.5, an open-source vision language model designed to interpret quantum processor diagnostics. This new version analyzes unfamiliar diagnostic results without prior training examples. What's interesting is that Ising Calibration 1.5 is 86.68% better than its predecessor when using examples from related experiments. It's also 11.4% smaller at BF16 precision, making deployment easier in local lab environments. For the first time, an NVFP4-quantized version is available, allowing deployment on a single GPU or an NVIDIA DGX Spark. This performance is comparable to leading closed models. The model is trained on data from various qubit modalities and evaluated using the QCalEval benchmark. The bottom line: This advancement helps automate quantum computer calibration and retuning operations.