This inventor holds 1 USPTO granted patent and 1 EPO patent. Top assignee: Nextvpu (Shanghai) Co., Ltd.. Active years: 2024.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2024
Title: Innovator Insights: Jun Pang and His Groundbreaking Patent
Introduction: Jun Pang, a notable inventor based in Shanghai, China, has made significant strides in the field of neural networks. With his innovative approach, he has contributed to the advancement of computing technologies through his patent focusing on memory pre-allocation.
Latest Patents: Jun Pang holds a patent for "Memory pre-allocation for forward calculation in a neural network." This invention presents a method that involves pre-allocating memory blocks for efficient processing in neural network computations. The method entails obtaining a memory block identifier from a list of idle memory blocks for each layer's output. If a suitable identifier is found, it is utilized; if not, a new identifier is created and stored in a mapping table, optimizing memory management and improving system performance.
Career Highlights: Currently, Jun Pang works at NextVPU (Shanghai) Co., Ltd., where he applies his expertise in developing advanced technological solutions. He has distinguished himself through his innovative contributions and is recognized as a key player in the field of neural networks.
Collaborations: In his endeavors, Jun collaborates with talented colleagues such as Haijiao Cai and Xinpeng Feng. Together, they contribute to the innovative environment at NextVPU, fostering collaboration that leads to significant advancements in technology.
Conclusion: Jun Pang exemplifies the spirit of innovation in the tech industry. His patent on memory pre-allocation for neural networks showcases his commitment to enhancing technological efficiencies. As he continues to work at NextVPU alongside his colleagues, there is no doubt that his contributions will further shape the landscape of artificial intelligence and computing technologies.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile