China’s national computing power network accepts first provinces as it moves to pool data centres to bolster infrastructure
- Data centres in southern Guangdong province and the southwestern provinces of Sichuan and Guizhou have been admitted to the network
- Demand for computing power is skyrocketing thanks to the rising popularity of artificial intelligence
China’s national computing power network has accepted its first batch of provinces as Beijing seeks to pool and build up national digital infrastructure despite increasingly limited access to advanced chips amid US-imposed trade restrictions.
Data centres in southern Guangdong province and the southwestern provinces of Sichuan and Guizhou have been admitted to the “China Computing Net” (C2NET), according to a report by government-backed newspaper Southern Daily that cited a Monday announcement at an industrial event in Shaoguan, Guangdong.
The C2NET was unveiled by the government last May to better consolidate and allocate computing power.
By pulling together regional resources, the network operates with a coordinated computing power of more than 3 exaflops, meaning the ability to perform 3 quintillion calculations per second, the report said.
As a reference, the world’s top supercomputer, the United States’ Frontier system, can achieve a maximal performance of 1.2 exaflops, according to ranking agency TOP500.
China’s internet watchdog warns of risks from generative AI, other advanced tech
Demand for computing power is skyrocketing thanks to the rising popularity of artificial intelligence (AI), and China is trying to ensure a sufficient supply of computing power to support the country’s research and technological progress.
“Computing power will become a type of utility like electricity that no matter where the supply is, one can pay to use it when there is demand for computing,” said Gao Wen, director of the Shenzhen-government-owned Peng Cheng Laboratory, which leads the C2NET project.
Gao added that computing power is also key to developing large language models (LLMs), the technology used to train the AI-powered chatbot ChatGPT. “The combination of computing power, data and algorithms is the core of LLMs,” said Gao, according to domestic media outlet Caijing.
Gao’s perspective is in line with comments by prominent Chinese venture capitalist Lee Kai-fu, who said over the weekend that one of the most promising AI investment strategies is computing infrastructure.
34:50
China makes first ChatGPT arrest as job losses by AI begin
Computing power has now become a key factor for a country’s technological growth, and the US has been restricting China’s access to advanced chips in recent times. Historian Chris Miller, the author of Chip War, has stated that “the rivalry between the United States and China may well be determined by computing power”.
For example, Nvidia’s data centre chip the A100 was added to the US export control list in August 2022, meaning it is unlikely China will have access to Nvidia’s GH200 super chip, its most powerful AI chip yet, which the chip maker said on Monday is now in full production.
Amid global interest in AI and the development of LLMs sparked by OpenAI’s ChatGPT, Chinese companies and research institutes including Baidu and Alibaba Group Holding have rushed to launch domestic alternatives. Alibaba owns the South China Morning Post.
Chinese institutions have so far launched at least 79 LLMs with over 1 billion parameters, a measure of the size and complexity of a model.
In comparison, OpenAI’s GPT-3.5 model has 175 billion parameters. Data has not yet been disclosed on the parameters for the updated GPT-4 model.
Fake or real? AI sparks fears for US 2024 White House race
The C2NET project is part of China’s ambitious “Eastern Data and Western Computing” plan to address regional imbalances in digital resources.
The project was announced in February 2022 by four central government agencies, under which data gathered from more prosperous areas in eastern China will be sent to poorer yet more spacious western areas for processing and storage.
Around 70 per cent of national demand for computing now comes from first-tier cities in China, located along the eastern and southern seaboard, Huang Hongbo, deputy general manager of CTYun, the cloud unit of state-owned telecoms operator China Telecom, said on Monday.
