The facility has become a key node in a national network of supercomputers. The test launch of the site took place in February of this year. At that time, the capacity was based on 30 thousand Chinese-developed AI accelerators. After modernization in April, the number of chips was increased to 60 thousand, which made the infrastructure the largest in the country in its segment.
An integrated ecosystem has been formed on the basis of the center, combining computing power, data arrays and modeling tools. Users have over 1,000 large language models with open weights at their disposal, ready for rapid deployment.
The technological feature of the platform is the automation of work processes. The system recognizes requests in natural language, independently distributes tasks and selects the necessary models. This approach allows us to provide results in a turnkey format, reducing the time required to conduct scientific research by a factor of several times.
The project provides end-to-end integration from general purpose solutions to specialized research. The cluster will become a reference model for coordinating distributed computing resources throughout the country.
The infrastructure operator intends to adhere to an open access policy, offering its services to both the academic community and the industrial sector. The launch of the facility is expected to significantly strengthen China’s position in global technology competition.
Previously we wrotethat China has launched an autonomous AI system based on a national network of supercomputers.
