Press Release

QuantaGrid servers improve AOI model precision and deployment speed with NVIDIA Defect Image Generation Skill

Taipei, Taiwan—July 20, 2026—Quanta Cloud Technology (QCT), a leading provider of AI and 5G solutions, today announced its collaboration with NVIDIA to integrate NVIDIA Defect Image Generation Skill, an agentic AI skill for synthetic data generation (SDG) powered by NVIDIA Cosmos, into QCT's server production lines. The collaboration enhances the precision and efficiency of industrial visual inspection, enabling faster and more accurate quality control.

In today’s server manufacturing sites, real defect data are scarce and take months to collect, label and iterate, creating a bottleneck for training and fine-tuning automated optical inspection (AOI) models. To shorten the time-to-online of high-accuracy AOI models, QCT worked with NVIDIA on an SDG proof of concept (PoC).

The PoC utilized a QuantaGrid server accelerated by NVIDIA accelerated computing, running NVIDIA Defect Image Generation Skill alongside QCT's existing AOI model. By providing a reference image of a defect-free product and text prompts describing specific anomaly types, the system generated synthetic videos containing multiple predefined defects, including Printed Circuit Board Assembly (PCBA) connector deformation, particle contamination, and failed components. These synthetic datasets were then used to train the AOI model for defect detection in real production environments.

The results demonstrated substantial improvements in both inspection performance and model training efficiency. After training with synthetic data, inspection accuracy increased by nearly 60%. In addition, the time required to collect defect data was reduced from up to one year using real-world samples to just one month using synthetic data—a reduction of more than 90%.

These improvements not only significantly reduced the labor overhead associated with manual inspection but also accelerated AI model deployment, helping manufacturers keep pace with the rapid product iteration cycles of modern AI infrastructure platforms.

“We are thrilled to integrate NVIDIA Defect Image Generation Skill, part of NVIDIA's Physical AI Data Factory Blueprint, into our manufacturing environment and take smart manufacturing to the next level,” said Mike Yang, President of QCT. “As a leading AI infrastructure provider, QCT not only delivers AI solutions but also applies AI to optimize its own operations. By deploying cutting-edge agentic AI in our factories, we are demonstrating how future manufacturing can leverage AI to achieve measurable business outcomes.”

Building on the success of this PoC, QCT plans to extend the deployment of NVIDIA Defect Image Generation Skill across its entire server manufacturing process, spanning component checking, surface mount technology (SMT) operations, and L10 system-level inspection. This collaboration marks an important milestone in QCT's smart manufacturing journey, demonstrating how agentic AI-generated synthetic data can unlock new levels of efficiency and quality across physical production environments.

To learn more, visit: https://blogs.nvidia.com/blog/siggraph-news-2026/#cosmos-3

About QCT

Quanta Cloud Technology (QCT) designs, manufactures, integrates, and services cutting-edge offerings for 5G Telco/Edge, AI/HPC, Cloud, and Enterprise infrastructure via its global network. Product lines include hyper-converged and software-defined data center solutions as well as servers, storage, and network switches from 1U to entire racks with a diverse ecosystem of hardware components and software partners to fit a variety of business verticals and workload parameters. www.qct.io

Other names and brands may be claimed as the property of others.

Media Contact

QCT Marketing Communication, marcom@qct.io