China's LineShine Supercomputer Reclaims World's Fastest Ranking
China's LineShine supercomputer has become the world's fastest system, surpassing the U.S. El Capitan by using a domestic CPU-only architecture to bypass American export controls.
The National Supercomputing Center in Shenzhen developed LineShine, which claimed the top spot in the June 2026 TOP500 rankings unveiled at the International Supercomputing Conference in Hamburg, Germany. LineShine achieved a sustained performance of 2.198 exaflops, becoming the first system to officially exceed 2 exaflops. It displaced the previous leader, the U.S. Department of Energy's El Capitan, which recorded 1.809 exaflops and fell to second place.
To circumvent U.S. export controls on advanced AI chips and GPUs, LineShine utilizes a unique CPU-only architecture. The system employs the LingKun platform with approximately 45,000 domestic LX2 processors and 13.79 million Armv9 cores designed by Huawei. While this design allows China to reach exascale performance without foreign hardware, the system is less energy-efficient than its competitors, consuming 42.2 megawatts compared to El Capitan's 29.7 megawatts.
Despite its raw speed in scientific benchmarks, LineShine ranked fourth in AI-specific tests, such as the HPL-MxP benchmark. This gap suggests a continued disparity in AI computing power compared to private U.S. systems. In response to China's progress, President Donald Trump signed an executive order to advance U.S. quantum computing capabilities. This marks the first time since 2017 that a Chinese supercomputer has led the global rankings, signaling a strategic push toward technological self-reliance.