On January 11th, the top 10 scientific and technological advancement of China and of the world in 2019 as voted upon by academics of the Chinese Academy of Sciences and the Chinese Academy of Engineering was announced in Beijing. “Hybrid-paradigm Tianjic Chip towards Artificial General Intelligence” by Shi Luping—professor of the Department of Precision Instrument at Tsinghua University—and his team was selected as one of the top ten scientific and technological advancements in China in 2019.
The Tianjic Chip is the first hybrid-paradigm brain-inspired computing chip for AGI. With a many-core architecture of computational memory fusion and decentralization, each chip contains 156 computational function cores, which is about 40000 neurons and tens of millions of synapses. The chip can not only support the independent deployment of computer-science-oriented machine learning algorithms and neuroscience-oriented neuromorphic computing models, but also support the heterogeneous modeling of the two, providing a computing platform for the development of AGI.
The first and second generation of the Tianjic Chip were respectively developed in 2015 and 2017. After a continuous improvement to its design, the second generation of the Tianjic Chip was endowed with high speed, high performance and low power consumption. The future development direction of the Tianjic Chip is to provide an energy efficienct, high speed and flexible computing platform for research into AGI. It can also be used for a variety of application development to promote research into AGI across different industries.
The paper Towards AGI with hybrid Tianjic chip architecture by Professor Shi Luping and his team was published as the cover article in the international journal Nature on August 1st, 2019.
（Source: Department of Precision Instrument, Tsinghua University）
http://www.dpi.tsinghua.edu.cn/publish/dpi/5114/2020/20200117160958147977772/20200117160958147977772_.html