Norma Proposes Lowering Quantum Training Costs With GPU-QPU Hybrid Use
AI TIMES ·
✦ AI Summary
According to AI TIMES, Norma said on the 9th that it can cut the real cost of using quantum hardware for training quantum models to about on…
According to AI TIMES, Norma said on the 9th that it can cut the real cost of using quantum hardware for training quantum models to about one-tenth by using software called 'Pado Pauli' that combines GPU and QPU. The method, unveiled at the International Quantum Industry Expo 2026, has the GPU handle most of the training computation and uses the QPU only for sampling at necessary stages. The company said it designed the software to reduce the computation load on actual hardware, targeting the particularly heavy time and cost burden involved in training quantum models. It also said it is operating a service that connects large language models to its quantum cloud to help with algorithm writing and execution. In new materials research, it is carrying out a proof of concept related to developing coating materials for solar cells using quantum generative models, and it is looking at ways to apply the technology in fields with limited data. On the commercialization side, it said it is focusing on what role quantum computing may play in improving AI, while also pursuing optimization of AI for small devices and development of genome analysis models.
Perspective
This case shows that the discussion around quantum computing is shifting beyond performance competition to how much it can actually lower usage fees and reshape research workflows. In particular, the approach of splitting roles instead of putting every computation on quantum hardware could lower adoption barriers and offer a practical path from research experiments to industrial applications. In the end, it suggests that the key to commercialization is moving from the hardware itself to cost efficiency and connectivity with existing AI environments.
This perspective is BizCrush's own commentary and is not part of the reporting by AI TIMES.
This article was produced with the help of an automated content generation algorithm.
Source: AI TIMES
View originalThis article was summarized and organized by BizCrush based on the original article from AI TIMES. For exact quotations and full details, please refer to the original article.