Chinese language Supercomputer Claims World High Spot, Exceeding 2 Exaflops Baseline
A newly unveiled Chinese language supercomputer has claimed the highest place on the authoritative TOP500 rating of the world’s strongest laptop programs, marking the primary time a Chinese language system has formally captured the number-one spot in practically a decade.
Introduced Tuesday on the ISC 2026 Excessive-Efficiency Computing convention, the system, named “LingSheng,” achieved a record-breaking sustained processing functionality of two.198 exaflops. The milestone makes it the primary supercomputer on this planet to formally breach the two exaflops efficiency ceiling on the standardized Excessive-Efficiency Linpack benchmark, working at over two quintillion floating-point operations per second.
Developed totally with home Chinese language parts and structure, the LingSheng system is deployed on the Nationwide Supercomputing Heart in Shenzhen. The machine represents a extreme break from conventional hybrid acceleration methodologies, using a pure central processing unit design constructed on China’s proprietary LingKun computing platform. The system is powered by 13.79 million unbiased computing cores consisting of 304-core LX2 processors working at 1.55 gigahertz, all tied collectively by a specialised customized interconnect known as LingQi and working on a home Kylin working system.
Past uncooked computational velocity, the platform secured the highest rating on the Excessive-Efficiency Conjugate Gradients check, an alternate metric designed to guage real-world engineering workloads and information entry speeds, whereas inserting fourth on the mixed-precision AI benchmarking check. The system attracts an estimated 42.2 megawatts of energy, yielding an power effectivity score of 52.07 gigaflops per watt, which lands it on the fiftieth spot on the accompanying Green500 power effectivity chart.
The up to date international index additionally featured a major new deployment from Western Europe. Italian power conglomerate Eni SpA efficiently launched its new HPC7 system, constructed by Hewlett Packard Enterprise Co., into the sixth spot on the worldwide chart. That European set up makes use of AMD Zen 4 EPYC processors alongside Intuition MI300A accelerated processing models to ship 571.50 petaflops of sustained efficiency, working alongside Eni’s older HPC6 system which presently occupies the eighth international place.
