NVIDIA Ising Enables Fully Automated Quantum Computer Calibration with Enhanced In-Context Learning

2026-07-27 · Source: NVIDIA Technical Blog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Emerging Technologies & Innovation · Depth: Advanced, short

Summary

NVIDIA has released Ising Calibration 1.5, an open-source 31-billion-parameter vision language model (VLM) designed for fully automated quantum computer calibration. This latest version enhances AI-based quantum processing unit (QPU) tuning by interpreting unfamiliar diagnostic results without prior training examples and leveraging related experiment data when available. Ising Calibration 1.5 is 11.4% smaller at BF16 precision, facilitating deployment in local lab environments. For the first time, an NVFP4-quantized version is available, enabling deployment on a single GPU or NVIDIA DGX Spark, comparable to leading closed models like Fable 5 and GPT 5.6 Sol. Trained on diverse qubit modalities, the model demonstrates strong performance on the QCalEval benchmark, outperforming other open models by 10% in zero-shot scenarios and showing an 86.68% improvement over its predecessor in in-context learning. NVIDIA provides full model weights, deployment recipes, and the QCalEval benchmark and dataset under an OpenMDW License.

Key takeaway

For quantum computing engineers tasked with automating QPU calibration, NVIDIA Ising Calibration 1.5 provides a robust, open-source vision language model. You can deploy its NVFP4-quantized version on a single GPU or DGX Spark, significantly reducing operational costs while achieving performance competitive with leading closed models. Consider integrating this VLM with the NVIDIA Nemo Agent Toolkit to streamline your calibration workflows and accelerate quantum experiment automation.

Key insights

NVIDIA's Ising Calibration 1.5 VLM automates quantum processor tuning, interpreting diagnostics without prior training and enhancing in-context learning.

Principles

• AI models can interpret complex diagnostic plots.
• In-context learning boosts calibration accuracy.
• Open-source VLMs rival closed, large-parameter models.

Method

Deploy agentic workflows using Ising Calibration 1.5 and the NVIDIA Nemo Agent Toolkit to automate quantum calibration experiments, interpreting diagnostic plots and recommending tuning steps.

In practice

• Deploy Ising Calibration 1.5 on a single GPU.
• Use NVFP4 quantization for efficient inference.
• Utilize QCalEval for VLM quantum calibration.

Topics

• Quantum Computing
• Vision Language Models
• QPU Calibration
• NVIDIA Ising Calibration
• In-Context Learning
• NVFP4 Quantization
• QCalEval Benchmark

Code references

• nvidia/QCalEval
• NVIDIA/Quantum-Calibration-Agent-Blueprint
• OpenMDW/OpenMDW

Best for: AI Scientist, Machine Learning Engineer, Research Scientist

Related on AIssential

• NVIDIA Ising Introduces AI-Powered Workflows to Build Fault-Tolerant Quantum Systems — NVIDIA Ising provides open AI models and tools to mitigate qubit noise and scale quantum computing.
• How AI Will Change Quantum Computing | NVIDIA AI Podcast Ep. 294 — AI models, like NVIDIA Ising, are critical for accelerating quantum computing by addressing key challenges such as error correction and calibration.
• An Analytically Trained Variational Surrogate for Quantum Phase Estimation on NISQ Hardware — A classically-trained shallow VQC can accurately surrogate Quantum Phase Estimation on NISQ hardware, bypassing exponential simulation.
• How AI Will Change Quantum Computing | NVIDIA AI Podcast Ep. 294 — NVIDIA I-Sing provides open AI models for quantum hardware calibration and error correction decoding.
• Graph Reinforcement Learning for Calibration-Aware Quantum Circuit Routing — Integrating real-time hardware calibration into quantum circuit routing via graph reinforcement learning significantly enhances circuit fidelity.
• NVIDIA's Quantum Day | here's a glimpse into the future... — NVIDIA's open Ising models accelerate quantum computing by integrating AI with GPU supercomputing for critical calibration and error correction tasks.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by NVIDIA Technical Blog.