China has reclaimed the top position in global high-performance computing with the launch of its domestically developed LineShine supercomputer, which has reportedly become the fastest system in the world.
The announcement was made at the ISC 2026 conference in Hamburg, Germany, where LineShine achieved a sustained performance of 2.198 exaflops, making it the first supercomputer to surpass the two-exaflop threshold in real-world sustained computing power.
The achievement marks China’s return to the top of global supercomputing rankings for the first time in nearly a decade.
Built in Shenzhen’s National Supercomputing Center
LineShine was developed at the National Supercomputing Center in Shenzhen, one of China’s leading hubs for advanced computing research.
The system has been designed to merge traditional scientific computing with modern artificial intelligence workloads, reflecting a broader global shift in how supercomputers are used. Rather than focusing solely on simulations, next-generation systems like LineShine are increasingly built to support AI-driven scientific discovery.
Powered by Domestic Processor Technology
At the core of the system are China’s self-developed LX2 processors, which are engineered to handle both high-performance computing and AI-based workloads.
These processors incorporate China’s first domestically produced high-bandwidth memory technology, significantly improving data transfer speeds—reportedly up to 10 times faster than conventional CPU memory systems.
This improvement is crucial in supercomputing environments, where bottlenecks in data movement can limit overall performance even when raw processing power is extremely high.
Massive-Scale Networking and Storage Architecture
Beyond processors, LineShine’s performance is driven by a large-scale interconnect system designed to efficiently link massive computing clusters.
The proprietary high-speed network can connect up to two million ports and 100,000 nodes, allowing the system to distribute workloads across an enormous computing infrastructure.
Its storage architecture is also designed for dual-purpose workloads, supporting both traditional simulation data and the large datasets required for AI training and inference.
Integrated Supercomputing and AI Platform
Developers have also built a unified software environment that combines supercomputing and artificial intelligence capabilities.
This integration allows researchers to run complex simulations and AI models within the same ecosystem, improving efficiency and reducing the time required for scientific discovery.
The system is also supported by a fully liquid-cooled design, which enhances energy efficiency while maintaining stable performance at extreme computational loads.
Applications Across Science and Industry
Since becoming operational, LineShine has already been used in a wide range of research fields, including:
Atmospheric and climate science
Ocean modelling
Engineering simulations
Advanced materials research
Drug discovery and pharmaceutical modelling
Brain science and neural research
AI-driven scientific computing
These applications highlight the increasing role of supercomputers not just in physics and engineering, but also in healthcare and artificial intelligence research.
A Shift in Global Supercomputing Power
Supercomputers are becoming central infrastructure for solving some of the world’s most complex problems, from climate change modelling to next-generation AI development.
LineShine’s debut signals a renewed shift in global high-performance computing leadership and underscores the rapid convergence between artificial intelligence and traditional supercomputing.
As AI continues to grow in importance, systems like LineShine suggest that the future of supercomputing will be defined not only by raw speed, but by how effectively machines can combine simulation, data processing and machine learning at unprecedented scale.
