China has reportedly switched on its “largest artificial intelligence computing cluster” in Zhengzhou, Henan province, on Tuesday, April 14. 

The South China Morning Post (SCMP) reported that it will be used for scientific research to accelerate drug discovery and find next-gen materials.

The cluster doubled its capacity in just two months, expanding from 30,000 to 60,000 AI accelerator chips.

Developed by Sugon (affiliated with the Chinese Academy of Sciences), the entire stack — from foundational chips and high-speed interconnects to the software — is produced within China to avoid external control.

CCTV reports that the Zhengzhou core node is now officially China’s most powerful platform for scientific AI computing.

Speeding up the future

Built for the AI for Science (AI4S) movement, this cluster moves beyond conversational AI to focus entirely on the labor of discovery.

The cluster is a fully self-reliant ecosystem, featuring domestic chips, networking, and software designed to tackle the most complex scientific puzzles. 

This total independence allows researchers to pursue their most sensitive work without the risk of foreign interference or technological blockades. 

In short, China has built a secure, high-performance sandbox where its scientific future is no longer tied to external suppliers.

At the national Changping Laboratory, the impact was immediate.

According to SCMP, the cluster has been advancing drug development by accelerating protein-folding simulations.

This massive speed boost turns biology’s toughest challenge — one that used to take a lifetime of work — into a task that can be finished in just a few days. The system could effectively clear the path for rapid medical breakthroughs.

The same story is playing out in materials science. Researchers are screening new elements in days rather than years, and aerospace engineers are simulating the turbulence of trillions of atmospheric grids simultaneously.

US-China race

To further lower the barrier to entry, Sugon launched OneScience, China’s first all-in-one development platform. 

SCMP reported that it hosts dozens of pre-loaded scientific models and datasets for conducting complex AI experiments without needing any background in computer programming.

It could open up world-class computing power to a much broader range of scientists in China.

To counter China’s progress, U.S. lawmakers recently proposed a bill to tighten global export controls on advanced chipmaking gear, aiming to align allies such as Japan and the Netherlands with American restrictions within 150 days. 

Moreover, the U.S. has launched the “Genesis Mission,” a Department of Energy initiative designed to build a unified AI platform across 17 national labs. 

This project aims to use federal data to train scientific models and deploy AI agents that can automate research and accelerate scientific development. 

China’s recent AI upgrade is a major strategic victory for Beijing, enabling researchers to overcome the constraints imposed by global GPU shortages and strict U.S. export bans. 

Using a system built entirely on domestic technology — ranging from homegrown silicon chips to specialized high-speed “scaleFabric” networking — the country has insulated its scientific progress from external geopolitical pressure. 

The shift from reliance on foreign tools to a fully sovereign infrastructure could help China pursue technological independence.

Recommended Articles