ShengJe Hung

Title: Innovations of ShengJe Hung

Introduction

ShengJe Hung is a notable inventor based in Hsinchu, Taiwan. He has made significant contributions to the field of machine learning and neural network computing. With a total of 3 patents, his work focuses on enhancing the performance of deep learning accelerators.

Latest Patents

One of his latest patents is titled "Instruction-based performance prediction for deep learning accelerator." This patent outlines a method for training a predictor that forecasts the performance of various machine learning models on different platforms. The method involves converting each model into a series of instructions and intermediate representations, simulating their execution, and generating performance reports. Another significant patent is "Neural network engine with tile-based execution." This invention describes an accelerator for neural network computing that includes hardware engines and buffer memory, designed to optimize convolution operations.

Career Highlights

ShengJe Hung is currently employed at MediaTek Corporation, a leading company in semiconductor technology. His work at MediaTek focuses on developing advanced solutions for machine learning and neural networks, contributing to the company's innovative product offerings.

Collaborations

He collaborates with talented coworkers such as Chien-Hung Lin and Meng-Hsuan Cheng, who share his passion for advancing technology in the field of machine learning.

Conclusion

ShengJe Hung's contributions to the field of deep learning and neural network computing are noteworthy. His innovative patents and work at MediaTek Corporation highlight his role as a key player in technological advancements.