Tsinghua University's heavyweight news: the world's first! Significant breakthroughs have been made in China's chip industry

Edited by: Huang ShengImagine a chip that integrates memory and computing capabilities, while protecting user privacy, and also has autonomous learning similar to the human brain. The energy consumption is only 1/35 of that of a specialized integrated circuit system under advanced technology

Edited by: Huang Sheng

Imagine a chip that integrates memory and computing capabilities, while protecting user privacy, and also has autonomous learning similar to the human brain. The energy consumption is only 1/35 of that of a specialized integrated circuit system under advanced technology. Does it sound magical?

According to the news on the official account of Tsinghua University on October 9, recently, professor Wu Huaqiang and associate professor Gao Bin of the School of Integrated Circuits of Tsinghua University based on the computing paradigm of integration of storage and computing,Developed the world's first fully integrated memristor memory computing integrated chip that supports efficient on-chip learning (machine learning can be directly completed on hardware), achieving significant breakthroughs in the field of memristor memory computing integrated chips that support on-chip learning, and is expected to promote the development of artificial intelligence, autonomous driving wearable devices, and other fields. The research findings were recently published in Science.

Memory resistor memory computing integrated learning chip and testing system (Image source: Tsinghua University)

Memory resistors are the fourth basic component of circuits after resistors, capacitors, and inductors. It can still "remember" the resistance state after a power outage and is used as a new type of nanoelectronic synaptic device.

Faced with the significant challenge of traditional memory and computing separation architecture restricting the improvement of computing power, Wu Huaqiang and Gao Bin are conducting research on the integrated memory and computing technology of focusing memory resistors to explore the implementation of a new paradigm for computer systems. The integrated memory and computing technology of memristors completely subverts von Neumann's traditional computing architecture in terms of underlying devices, circuit architecture, and computing paradigms, achieving leapfrog improvements in computing power and energy efficiency. At the same time, this technology can also utilize the learning characteristics of underlying devices to support real-time on-chip learning and empower edge training new scenarios based on local learning.

The research group creatively proposes a new universal algorithm and architecture for efficient on-chip learning through the integration of memory and computing, based on the paradigm of integrated storage and computing. Through collaborative innovation throughout the entire process of algorithm, architecture, and integration methods, the world's first fully integrated memory and computing integrated chip that supports efficient on-chip learning has been developed.

The development of memristor chips faces both technical and engineering challenges.

It is understood that the development of memristor chips involves cutting-edge knowledge in multiple disciplines such as materials science, physics, and electronic engineering. Among many technical challenges,The first issue to be addressed is how to achieve large-scale integration of memristor devices. Through extensive experiments and theoretical research, the team proposed an architecture circuit process collaborative optimization method, providing guidance for the design of an integrated storage and computing system.

With the process of large-scale integration and key circuit design,How to overcome the errors caused by low-level multi-scale non ideal and assemble them into an efficient system chip?With the joint efforts of team teachers and students, the STELLAR architecture has been studied and proposed, algorithm optimization and simulation experiments have been completed, and a highly efficient integrated memory computing integrated learning chip has been prepared, achieving significant improvements in speed and energy efficiency.

Looking ahead,Wu Huaqiang hopes that the team's solutions and technologies can go out of the laboratory, effectively promote the transformation of scientific research achievements, and be committed to serving the needs of the country and society.

Daily Economic News Comprehensive Tsinghua University official account

Daily Economic News

Disclaimer: The content of this article is sourced from the internet. The copyright of the text, images, and other materials belongs to the original author. The platform reprints the materials for the purpose of conveying more information. The content of the article is for reference and learning only, and should not be used for commercial purposes. If it infringes on your legitimate rights and interests, please contact us promptly and we will handle it as soon as possible! We respect copyright and are committed to protecting it. Thank you for sharing.(Email:[email protected])