Sejong University Advances Next-Generation GPU Research
Sejong University said Monday that its Future GPU Research Institute has accelerated the development of core technologies for upcoming graphics processing units (GPUs) and artificial intelligence (AI) semiconductors. The university featured the institute’s work on technologies designed to support future advances in computer graphics and AI hardware.
The institute is led by Professor Park Woo-chan from the Department of Computer Science and Engineering. Its current research focuses on real-time path tracing, regarded as the next stage in GPU rendering beyond conventional ray tracing. Ray tracing is a computer graphics technique that creates highly realistic images by simulating how light moves and interacts with objects in a scene. Path tracing builds on this approach by calculating the complete path of light as it reflects throughout a scene. This process produces images that closely resemble real-world lighting conditions.
Although path tracing can deliver highly realistic graphics, the technology has encountered considerable barriers to wider commercial use. One of the main challenges has been the considerable visual noise produced during instantaneous rendering, which makes feasible implementation difficult in many applications.
To overcome this limitation, the Future GPU Research Institute is developing its own real-time AI denoising semiconductor technology. According to the university, Professor Park’s research team has established a substantial technological foundation through several years of research. In addition, the team has filed and secured dozens of patents, showing its work in this area of semiconductor development.
The university stated that the institute’s AI microchip technology differs from conventional GPUs and neural processing units (NPUs). The technology is designed to deliver instant processing and perform AI analyses without requiring access to external memory. This approach enables a lightweight and efficient hardware implementation.
The university also reported that the technology could be applied across a range of industries. Potential areas of use include surveillance camera systems, self-driving vehicles, medical imaging and defence-related applications.
Alongside its work on GPU technologies, the institute has broadened its research activities to include audio AI integrated-circuit technology. Operating under the university’s Industry-Academy Cooperation Foundation, the institute is developing technologies for applications such as hearing aids and AI-enabled smart glasses.
The university stated that the institute continues to expand collaborations with research organisations and industry partners, both within South Korea and internationally, while pursuing the commercialisation of technologies developed through its patented research.
Professor Park stated that the institute is still focused on promoting research in next-generation GPU technology, an area in which major international technology companies continue to invest heavily in R&D. He also stated that the institute is working to advance its technologies for commercial application while continuing research to improve its position within next-generation GPUs and AI semiconductors.
The university identified Professor Park as an internationally recognised specialist in GPU ray-tracing technology and noted his long-standing role in GPU research in South Korea. His work has contributed to the development of ray tracing, a graphics technology widely used across gaming, filmmaking and virtual reality applications.
The university also noted that demand for photorealistic graphics has continued to increase, resulting in stronger competition in GPU development. As a result, most high-performance and mobile GPUs released by leading technology companies, including Nvidia, AMD, Intel, Apple, Qualcomm and Samsung Electronics, now support real-time ray tracing.
In addition to leading the institute’s research activities, Professor Park is conducting two projects funded by the Ministry of Science and ICT. One project is centred on developing ultrarealistic extended reality technologies for a metaverse that connects the physical and virtual worlds. The second project focuses on strengthening intellectual property and advancing the commercialisation of real-time AI denoising hardware technology.
The university stated that these projects form part of the institute’s broader research programme in GPU and AI semiconductor technologies. Its operations continue to cover image rendering, AI hardware, and related semiconductor technologies, while supporting research for real-world applications across multiple sectors.
