IonQ, a quantum computing leader based in College Park, MD, has unveiled significant advancements in hybrid quantum-classical approaches for artificial intelligence and machine learning applications. Through collaborations with a major automotive manufacturer and the AIST Global Research and Development Center for Business by Quantum AI (G-QuAT), IonQ demonstrated how integrating quantum computing can address data scarcity challenges in fields such as materials science and large language models. The company used quantum-enhanced generative adversarial networks (QGANs) to produce synthetic images of rare microstructure anomalies, and applied quantum machine learning (QML) to refine and optimize large language models (LLMs).
IonQ´s hybrid quantum-classical architecture enables new possibilities for LLM fine-tuning. By embedding a parameterized quantum circuit as an added layer within open-source LLMs, the system achieved enhanced classification accuracy compared to purely classical methods. Testing revealed that model accuracy improved as the number of qubits increased, with additional benefits of lower energy consumption on complex problems. These results are especially relevant in applications where high-performance improvements and energy efficiency are critical, illustrating IonQ’s drive to push the boundaries of Quantum Machine Learning for next-generation artificial intelligence systems.
In the realm of materials science, IonQ´s QGAN technology allowed for the generation of synthetic steel microstructure images using a trapped-ion quantum computer. This created augmented datasets that significantly improved the diversity and quality of training data for industrial artificial intelligence systems, resulting in higher image quality scores in 70% of cases versus classical approaches. The research offers practical solutions to longstanding challenges in industrial AI, such as limited or imbalanced datasets, and underscores the promise of quantum computing in optimizing materials characterization and manufacturing workflows. IonQ’s Forte Enterprise systems, offering 36 algorithmic qubits and accessible via major cloud providers, support these initiatives and deepen the company’s role in advancing quantum technologies for both commercial and research users. Strategic partnerships, including the agreement with AIST’s G-QuAT, further highlight IonQ’s commitment to developing powerful quantum-classical hybrid technologies for real-world problem-solving.
