Chao-Yang Kao

Title: Chao-Yang Kao: Innovator in Neural Network and Matrix Calculation Technologies

Introduction

Chao-Yang Kao is a prominent inventor based in Hsinchu, Taiwan. He has made significant contributions to the fields of neural networks and matrix calculations, holding a total of 10 patents. His innovative work has advanced technology in various applications, particularly in computational methods.

Latest Patents

Among his latest patents is a harmonic densely connecting method of a block of convolutional neural network model and the system thereof. This method includes an input step, several layer operation steps, and an output step. The input step involves storing an original input tensor into memory. Each layer operation step consists of a layer-input tensor concatenating step and a convolution operation step. The concatenating step selects layer-input element tensors based on an input connection rule, while the convolution operation calculates a result tensor to be stored in memory. The output step then produces a block output.

Another notable patent is the matrix calculation device and its operation method. This device includes a storing unit, a multiply-accumulate (MAC) circuit, a pre-fetch circuit, and a control circuit. The storing unit holds two matrices, while the MAC circuit executes calculations. The pre-fetch circuit retrieves data from the storing unit to optimize the MAC calculations, enhancing the efficiency of matrix operations.

Career Highlights

Chao-Yang Kao has worked with notable companies such as Neuchips Corporation and Novatek Microelectronics Corporation. His experience in these organizations has allowed him to apply his innovative ideas in practical settings, contributing to advancements in technology.

Collaborations

He has collaborated with esteemed colleagues, including Youn-Long Lin and Chiung-Liang Lin. Their teamwork has fostered an environment of innovation and creativity, leading to the development of groundbreaking technologies.

Conclusion

Chao-Yang Kao's contributions to neural network and matrix calculation technologies exemplify his role as a leading inventor. His patents reflect a commitment to advancing computational methods, making a lasting impact on the field.