Thanks to the innovation of cloud computing and storage technology, the development of modern science has also ushered in new opportunities. However, while promoting the innovation of scientific research work, the scientific research platform needs to solve many problems in data storage and capacity calculation at the same time. Taking the University of Science and Technology of China (hereinafter referred to as USTC) as an example, the overall resource utilization rate of the existing computing centers on the campus is close to the limit, and new solutions are urgently needed. For this, Inspur Information provides a more flexible and reliable distribution storage solution.
During the period, the overall plan centered on "computing + storage" upgraded the cluster system. China University of Science and Technology planned and deployed 25 Inspur information GPU server nodes and a set of Inspur information distributed parallel file storage AS13000 series to provide calculation and data for the scientific research platform. Strong infrastructure support.
Aggregate read and write bandwidth performance
The data storage cluster has a high redundancy mode, that is, when a node fails, it can be automatically switched to a backup node, the frequency of equipment failures is reduced, and the data of the scientific research platform is not lost, ensuring data security and continuity of project operation. In addition, the University of Science and Technology of China considers the increase in the future research direction and scale of the scientific research platform, and the Inspur Information distributed parallel file storage AS13000 can expand new nodes online according to the growth of later business, and the performance and capacity of the system will increase with the number of nodes. The increase increases linearly, which protects the continuous research of scientific research projects.
Logical Topology Diagram of University-level Computing Platform of China University of Science and Technology
The university-level computing platform of the University of Science and Technology of China, which was officially established in 2003, is one of the six public experimental centers of the school and one of the earliest platforms in domestic universities. At present, the scale of the computing cluster of the University of Science and Technology of China exceeds 1,000 units, and the computing capacity reaches 5 petaflops per second. Support the publication of 464 SCI papers in 2022 (including 38 papers in top journals such as Nature series, Science series, PRL, JACS, Cell, etc.), provide data and computing platforms for technological innovation in frontier fields, and play a role in achieving remarkable project results for users Significant support.
The pace of scientific and technological innovation continues, and the promotion of innovation in scientific research requires a more powerful computing platform as a driver. As a professional service provider in the storage field, Inspur Information can provide a better solution for major scientific research platforms and assist the platform to efficiently cope with new technologies. A round of computing and data resource storage problems, thus laying a solid foundation for accelerating cutting-edge scientific and technological research.