Search
Nature Communications | New Progress Achieved by Wang Zhongrui's Research Group from the Shenzhen-Hong Kong Microelectronics Institute at the Southern University of Science and Technology in Memristor
4 hour ago /
Read about 0 minute
Author：小编
Generative artificial intelligence (AIGC) is developing rapidly in areas such as text-to-image, text-to-video, embodied intelligence, and world models. Generative algorithms, represented by diffusion models, are becoming key technologies for simulating human imagination. However, diffusion models face issues of slow speed and high energy consumption on traditional digital computing platforms. The reason lies in the separation of memory and computing units in the von Neumann architecture, which leads to massive data movement; at the same time, diffusion models rely on solving neural differential equations, and digital computers need to discretize the continuous generation process, increasing time and energy consumption costs.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
