China has taken a major step in the global supercomputer race after LineShine climbed to the top of the TOP500 ranking. The system has overtaken El Capitan from the United States, ending America’s hold on the fastest-supercomputer title.
The shift matters well beyond prestige. Supercomputers support medical research, climate modeling, aerospace engineering, fusion studies, and other heavy computing workloads that influence national competitiveness.
A different kind of top-tier machine
LineShine recorded 2.198 Exaflop/s on the standard HPL benchmark, according to TOP500. That result is about 21.5% higher than El Capitan’s 1.809 Exaflop/s.
TOP500 released its latest ranking on 23 June 2026, and LineShine took first place in the list. It is the first time since 2017 that the United States has lost the fastest-supercomputer crown.
One of the most notable aspects of the machine is its scale. LineShine is said to be about 11 million times faster than a typical home or office PC.
Built entirely with domestic chips
LineShine was developed by the National Supercomputing Centre in Shenzhen, or NSCC-SZ, together with the Shenzhen Cloud Computing Center. The system was designed to showcase a fully domestic hardware stack.
Unlike many leading supercomputers that still depend on Intel, AMD, or other major foreign suppliers, LineShine uses homegrown chips. NSCC-SZ built the machine around LX2, also known as LingKun, a processor based on the ARM architecture.
The system also carries an enormous processing footprint. LineShine has around 13.79 million cores, which helps explain its leap in benchmark performance.
Why the architecture stands out
Most modern supercomputers rely on a hybrid CPU-plus-GPU design, and El Capitan fits that model. LineShine takes a different route by reaching peak performance with CPUs only.
That choice gives the system both technical and strategic value. China is not only fielding a faster machine, but doing so with an architecture that breaks from the dominant pattern in high-end computing.
A stronger position in the technology race
LineShine is being viewed as an important milestone for China’s high-performance computing, or HPC, independence. The achievement shows that a world-class system can be built from domestic components even under intense technology pressure.
In recent years, China’s access to HPC chips from Intel, AMD, and NVIDIA has been restricted by U.S. sanctions and export controls. LineShine reduces that dependency and strengthens China’s position if further restrictions emerge.
From a broader geopolitical perspective, the message is clear. The supercomputer race is now about more than raw speed, because chip independence and system autonomy have become part of the competition.
What still remains unresolved
Even with this breakthrough, China still faces bigger challenges. Building a complete domestic computing chain, from chips to software and systems, remains a larger goal.
There are also continuing gaps in AI hardware. Semiconductor self-sufficiency and the ability to scale AI computing infrastructure remain key tasks for the country.
For researchers and industry, the change at the top of TOP500 is worth watching closely. The country leading supercomputers usually gains an advantage in scientific simulation and other work that depends on extreme computing power.
LineShine does more than move China ahead in one ranking. It shows that the world’s fastest supercomputer can be built with local chips, a CPU-only design, and enough scale to reset expectations in global computing.
