Meng Li

Title: Meng Li: Innovator in Neural Network Technologies

Introduction

Meng Li is a prominent inventor based in Union City, CA (US), known for his contributions to the field of neural network technologies. With a total of 5 patents, he has made significant advancements that enhance the efficiency and functionality of neural network accelerators.

Latest Patents

One of Meng Li's latest patents is titled "Systems and methods for reading and writing sparse data in a neural network accelerator." This invention includes a system, method, and device for efficiently managing sparse data within a neural network accelerator. It utilizes a mask to identify byte positions in a data word that contain non-zero values, allowing for optimized data storage and retrieval. Another notable patent is "System and method for performing small channel count convolutions in energy-efficient input operand stationary accelerator." This invention focuses on generating outputs for neural network layers by partitioning input data into subarrays, which are then processed by a group of processing elements (PEs) to enhance computational efficiency.

Career Highlights

Meng Li has worked with leading technology companies, including Meta Platforms Technologies, LLC and Facebook Technologies, LLC. His experience in these organizations has allowed him to collaborate on cutting-edge projects that push the boundaries of neural network applications.

Collaborations

Throughout his career, Meng has collaborated with talented individuals such as Ganesh Venkatesh and Liangzhen Lai. These partnerships have contributed to the innovative solutions he has developed in the field of neural networks.

Conclusion

Meng Li's work in neural network technologies exemplifies the impact of innovation in the tech industry. His patents reflect a commitment to advancing the capabilities of neural networks, making significant contributions to the field.