Foreign Media: China's "AI Research" Computing Power Doubles in Two Months, Entirely Using Domestic Chips
China's largest scientific AI computing cluster has officially gone online recently, deployed at the core node of the Henan Zhengzhou National Supercomputing Network. The AI acceleration cards used in this cluster are entirely independently developed and manufactured by Sugon, a company under the Chinese Academy of Sciences. Since early February this year, during just two months of trial operation, the number of chips has increased from 30,000 to 60,000, doubling the computing power.
The Zhengzhou core node has now become China's most powerful scientific intelligent computing infrastructure, considered a significant breakthrough in China's effort to "secure a leading position" in AI-driven scientific research.
Chinese researchers have long faced bottlenecks such as insufficient computing power, software restrictions, and reliance on foreign suppliers. This recent expansion marks a crucial step forward in China’s efforts to overcome external dependencies and build an independent, controllable AI research infrastructure.
Original Article: toutiao.com/article/1862610000319560/
Disclaimer: The views expressed in this article are solely those of the author.