IEEE, the global champion of the international hardweight competition, FPGA, the immediate 3D scenario technology that beats the international famous school.

The team of the Institute of Electronics of the National Yang Ming University of Transport has achieved significant results in the area of international reorganisation of computing. Four master ' s students, Rygongwei, Hu Xiaomun, Guo Jia Woo and Zheng Xiaoqing, under the guidance of Assistant Professor Wu Yi Zhong, won the world ' s first place at **FCCM 2026 Reconfigurable Company Challenge (RCC)** in Atlanta, United States.

According to Ying-ming, the team's 16 May 2026 First place in the competition. A High-Throughput FPGA Accelerator for Real-Time 3D Gaussian Splatting 。 The follow-up news from Yang Ming University added that the team had defeated international school teams, including Cornell University, Carnegie Mellon University, San Diego University of California and Beijing University.

Research core: making 3D Gaussian Spratting faster and more energy-efficient

Team focus on the rapid development of recent years 3D Gaussian Splatting（3DGS） Technology. Such methods can be used for immediate 3D scene reconstruction and visualization, but the computation process involves a large amount of information and depth sequencing, which places a significant burden on immediate displays.

The approach of the Ying-ming team is to design high-volume structures that can be implemented on the FPGA platform, both from algorithms and hard-body structures, to lower the ranking and operating burden and to increase the efficiency of the immediate landscape reconstruction.

The FCCM 2026 paper input from the Yingming High R & D Platform further shows that the work was done on AMD-Xilinx Alveo U200 FPGA and reached at 600x340 resolution 637.6 FPS It's about time. 2.7 W ； Compared to previous FPGA-based creating energy efficiency improvements 7.64 to 12.76 。

It's not just a competition, but it's in a collection of IEE papers.

This outcome is not confined to the competition presentation. ** The collection of papers by INEE 34th Annual International Symposium on Field-Programable Custodial Construction Machines (FCCM 2026)**, pp. 285-288, DOI 10.1109/FCCM68464.2026.00073。

FCM 2026 was held in Atlanta, United States, from 13 to 16 May 2026. The FCM long-term focus on the FPGA, reorganizable computing, efficient computing and hardweight acceleration are important international seminars for the relevant research communities.

Meaning for Taiwan students and higher education

This result has two interesting orientations for students in Taiwan who are prepared for electric power, electronics, information engineering, artificial intelligence or chip design.

First, international competitions are not separate from formal academic research. The team completed the system at the same time, presented the competition and published thesis, reflecting the growing emphasis on the integrity of high-level engineering education, from algorithms, hardware and software to system integration.

Second, the FPGA, AI accessation, 3D scene review are meeting quickly. Such cross-cutting hardware acceleration capabilities would be of considerable research and industry value if students were interested in future AI chip, computer visualization, AR/VR, self-driving systems or robotics.

On July 14, 2026, the day and day media also reported the results. Prior to its publication, the newspaper had cross-checked the names of students, the names of competitions, the timing of the award, the subject of research and the information on dissertations with the Yang-ming Research and Development Service, the official news and research platform of the school.