China Telecom's Advancements in Computing Infrastructure and AI Research

TapTechNews July 4th news, Ke Ruiwen, the chairman of China Telecom Group Co., Ltd., said at the 2024 World Artificial Intelligence Conference held today that China Telecom moderately advanced the construction of intelligent computing infrastructure and promoted the upgrade of data centers to AIDC.

For the Yangtze River Delta region with strong demand, he said that China Telecom also recently launched the computing power dispatching center in the Yangtze River Delta to realize the intensive supply and unified dispatch of computing power resources.

In May this year, Shao Guanglu, the general manager of China Telecom, once disclosed that China Telecom focused on planning the intelligent computing layout of 2+3+7+N+M, and the intelligent computing scale reached 13 EFLOPS (TapTechNews note: 100 gigabytes per second, that is, 10^18 floating-point operations).

It is said that China Telecom has built a single-cluster 10,000-card intelligent computing resource pool in Shanghai; in terms of intelligent computing cloud capabilities, it has initially possessed the three major ability platforms of Yunxiao, Huiju and Xiran, realizing the abilities of unified management, one-stop provision and heterogeneous dispatch of intelligent computing power, and the intelligent computing cloud capabilities have initially taken shape.

In June this year, the China Telecom Artificial Intelligence Research Institute (TeleAI) jointly released the world's first single-dense trillion-parameter semantic model Tele-FLM-1T with the Beijing Academy of Artificial Intelligence, becoming one of the first institutions in China to release a dense trillion-parameter large model.