Shulou(Shulou.com)11/24 Report--

Thanks to CTOnews.com netizens for the delivery of clues on the way! CTOnews.com, May 21, May 18, 2023 the Seventh World Intelligence Congress officially began in National Exhibition and Convention Center (Tianjin), with a four-day exhibition period of intelligent technology.

The intelligent science and technology exhibition has planned eight pavilions and outdoor exhibition areas. there are thematic exhibition areas such as information technology application innovation, artificial intelligence, 5G + industrial Internet, intelligent transportation, intelligent manufacturing, smart life, digital finance, digital health, international and provincial cities, as well as intelligent driving and 5G experience areas.

Tianhe new generation supercomputer is a new generation of 10 billion supercomputers in China, and it also made its debut at this exhibition. Over the years, the Tianhe supercomputer has gone from trillions to billions of times (CTOnews.com Note: the new generation supercomputer is not Tianhe-1, its double-precision floating-point peak computing performance is 200p Flops, the data storage capacity is not lower than 20PB, and the peak power consumption is not higher than 8MW, which is 150 times higher than the computing power of "Tianhe-1"), constantly challenging the speed limit of the world's computing power. Tianjin, relying on Tianhe, accounts for more than 20% of the country's supercomputing resources.

In order to give full play to the advantages of supercomputing in computing power and promote the application of domestic heterogeneous supercomputer platforms in the field of AI application development and services, the National supercomputing Tianjin Center released "Tianhe E-level Intelligent Computing Open Innovation platform" and "Tianhe Tianyuan Model" at the World Intelligent Science and Technology Innovation Cooperation Summit held during the Seventh World Intelligent Congress.

▲ source: national supercomputing Tianjin Center it is introduced that the data set of the Chinese big language model is very scarce. The National Supercomputing Tianjin Center collates web data, various open source training data, Chinese novel data, ancient Chinese data, encyclopedic data, news data, as well as professional TCM, medicine, consultation, law and other data sets, with a total token of 350B, and finally trained to create its own Chinese language model-Tianhe Tianyuan model.

The new generation of Tianhe supercomputer has realized comprehensive and independent information technology application innovation from hardware to software environment, and officials have also demonstrated in the exhibition area the core technologies such as computing chips and motherboards used in Tianhe new generation supercomputers.

It is reported that the "Tianhe E-level Intelligent Computing Open Innovation platform" will bring cross-modal super computing power to break through 10 billion times, support traditional scientific and engineering computing, and serve intelligent hybrid computing. To create an omni-directional computing power enabling innovation and high-quality development carrier of the digital economy.

In the future, the National Supercomputing Tianjin Center will also start in-depth training of professional models in medical, industrial, legal and other fields on the basis of the "Tianhe Tianyuan Model". In the future, the large model will become the intelligent base of the industry under the support of deep learning platform.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.