China’s LineShine supercomputer reclaims top spot on TOP500 with 2.198 exaflops
China’s LineShine supercomputer tops the TOP500 rankings, achieving 2.198 exaflops and becoming the world’s fastest known system for the first time since 2017. (top500.org)
Lead: LineShine debuts at No. 1 on TOP500
LineShine’s placement at the top of the June 2026 TOP500 list marks a major shift in the global supercomputing landscape and returns China to the highest position last held in 2017. The system posted 2.198 exaflops on the High Performance Linpack (HPL) benchmark, making it the first TOP500 system to exceed 2 exaflops on sustained FP64 performance. (top500.org)
The machine is installed at the National Supercomputing Centre in Shenzhen and was presented at the International Supercomputing Conference (ISC) in Hamburg. The announcement ends the recent run of the United States’ El Capitan at the summit of the list. (apnews.com)
Ranking change ends El Capitan’s run
The El Capitan system at Lawrence Livermore National Laboratory, which measured 1.809 exaflops on the same benchmark, moved to second place after LineShine’s debut. The margin between the two systems is roughly 20 percent based on HPL measurements. (tomshardware.com)
El Capitan, Frontier and Aurora remain among the highest-ranked systems, keeping the United States well represented in the upper ranks even as China reclaims the top position. Observers say the reshuffle highlights a fast-moving period of investment and deployment in high-performance computing. (fortune.com)
System location and builders
LineShine is housed at the National Supercomputing Centre, Shenzhen (NSCS), and the build and deployment were managed by the Shenzhen Cloud Computing Center. The system’s architecture relies on a vast array of Arm-based cores and is reported to use a wholly domestic technology stack. (top500.org)
Officials and technical briefings at ISC indicated the machine was tested with 13,789,440 cores to achieve the measured HPL result, underscoring the scale of the installation and the density of the underlying compute fabric. (top500.org)
Technical profile and domestic components
LineShine’s debut performance was notable because it reached exascale-class sustained double-precision performance using a CPU-only design, without the widespread deployment of GPUs that characterizes many recent exascale systems. This places emphasis on China’s ability to scale CPU-based architectures at very large node counts. (tomshardware.com)
The system’s designers have stressed the use of domestically produced processors and interconnects, aligning with national strategies to reduce dependency on foreign chips and to secure supply chains for strategic computing infrastructure. Analysts say this emphasis on a CPU-only approach differentiates LineShine from some competing systems that rely on GPU accelerators. (supercomputing.news)
Implications for scientific research and industry
Supercomputers such as LineShine are central to climate modelling, materials science, industrial research and advanced AI workloads, and a higher TOP500 ranking can accelerate collaborations and project allocations. Institutions that secure access to top-ranked systems can run larger simulations and tighter code-to-hardware co-design cycles. (networkworld.com)
Beyond pure science, national prestige and strategic competition factor into investments in supercomputing, with governments citing scientific, economic and security returns when funding major HPC facilities. The LineShine announcement is therefore likely to renew focus on capacity-building programs worldwide. (fortune.com)
TOP500 process and ISC announcement
The TOP500 list, compiled biannually on the basis of HPL benchmark results, was unveiled at ISC 2026 in Hamburg, where vendors and national teams presented system details and baseline measurements. The June listing is the 67th edition and reflects system submissions validated by the TOP500 project. (top500.org)
Presenters at ISC noted that the HPL benchmark remains an imperfect but standardised yardstick for peak sustained floating-point performance, and TOP500 maintainers continue to add contextual metrics such as energy efficiency and alternative benchmarks to capture broader system capability. (top500.org)
China’s return to No. 1 on the TOP500 list with the LineShine system is likely to spur additional deployments and upgrades across national supercomputing programmes, while also prompting a reassessment of procurement and research priorities among competitors. The new ranking underscores both the rapid pace of technological development in high-performance computing and the strategic value nations place on computational leadership.