Mirror Review
June 24, 2026
China’s LineShine is the new World’s Fastest Supercomputer after debuting at number one on the 67th edition of the Top500 list. Announced at the ISC 2026 conference in Hamburg, Germany, this system officially ends a seven-year drought for China at the top of global supercomputing rankings.
China’s newly unveiled supercomputer achieved an unprecedented 2.198 Exaflops on the High Performance Linpack (HPL) benchmark, surpassing the previous leader, El Capitan, from the United States.
This win proves that domestic innovation can thrive independently of foreign hardware restrictions.
China’s Return to the Top of Global Supercomputing
The arrival of the supercomputer LineShine changes the balance of power in high-performance computing.
China last held the top spot in 2017 with the Sunway TaihuLight system. In the years that followed, Chinese laboratories became increasingly quiet, choosing not to submit their benchmark results to the Top500 list amid growing trade tensions and export controls.
According to reports, China’s LineShine was built without public government funding. Because it was privately funded, its designers felt free to submit its data for official global benchmarking. The results shocked the industry by beating El Capitan, the Lawrence Livermore National Laboratory system that had held the crown since November 2024. LineShine outperformed El Capitan by more than 20%, carrying out over 2 quintillion calculations per second.
Technical Specifications of the World’s Fastest Supercomputer LineShine
What makes LineShine unique is its internal design.
Most modern exascale systems rely heavily on Graphics Processing Units (GPUs) to handle massive computational loads. LineShine completely abandons this traditional layout. Instead, it runs entirely on general-purpose Central Processing Units (CPUs) equipped with custom built-in acceleration.
Here are the core technical facts behind this massive machine:
- Processor Architecture: Custom Chinese 304-core LX2 processors running at 1.55 GHz.
- Total Core Count: 13.79 million computing cores spread across 90 hardware cabinets.
- Platform and Interconnect: Built on the “LingKun” platform and linked by the proprietary LingQi interconnect.
- Operating System: Kylin OS.
- Power Consumption: Draws 42.2 megawatts of power, reaching an energy efficiency of 52.07 Gigaflops/Watt.
LineShine is the first and only system to break the two-exaflop barrier using a CPU-only design. Its processors use an original design based on an instruction set licensed from Arm Holdings.
Overcoming the GPU Shortage Through Smart Engineering
The CPU-only layout of LineShine represents a major engineering win.
The United States government has placed strict export controls on advanced GPUs to limit China’s progress in artificial intelligence and high-performance computing. By building a world-class system without these restricted chips, Chinese scientists proved they could adapt to international roadblocks.
Jack Dongarra, a founder of the Top500 list and a computer science professor at the University of Tennessee, recently inspected the new Chinese machine. He praised the achievement, noting its independence from Western supply lines.
“It’s an impressive system,” Dr. Dongarra stated. “They upped us by developing a system that is not reliant on GPUs.”
He also noted that trade limits forced China to spend heavily on its own hardware designs, which led directly to this breakthrough.
Speaking at the Hamburg conference, LineShine chief designer Lu Yutong explained that the machine features specialized circuitry to handle matrix and vector calculations directly on the CPU, allowing it to manage traditional science tasks and AI training simultaneously.
What is LineShine Being Used For?
Supercomputers are not just built for breaking records; they are vital tools for national security, medicine, and industrial production. The Shenzhen center confirmed that LineShine completed its full deployment by the end of 2025 and is already working on complex assignments.
The machine currently runs a sophisticated simulation of Earth that tracks how the atmosphere, oceans, land, and ice interact. It also processes intricate neural mapping models for neuroscience research, assists in drug discovery projects, and runs complex engineering simulations.
The design team has also submitted 14 papers to the prestigious Gordon Bell Prize, with three projects named as finalists for the core award and three others competing for the climate science prize.
Inside the June 2026 Top500 Supercomputer Rankings
The latest Top500 supercomputer rankings show a high competition for the title of the world’s fastest supercomputer. Five systems have now crossed the elite exascale threshold, distributing world-class computing power across Asia, North America, and Europe simultaneously.
The Top 10 Most Powerful Supercomputers in the World
| System Name | Installation Site / Country | HPL Performance (Exaflop/s) |
| 1. LineShine | National Supercomputing Centre, Shenzhen / China | 2.198 |
| 2. El Capitan | Lawrence Livermore National Laboratory / United States | 1.809 |
| 3. Frontier | Oak Ridge National Laboratory / United States | 1.353 |
| 4. Aurora | Argonne National Laboratory / United States | 1.012 |
| 5. JUPITER Booster | Jülich Supercomputing Centre / Germany | 1.000 |
| 6. HPC7 | Eni S.p.A. / Italy | 0.571 |
| 7. Eagle | Microsoft Azure / United States | 0.561 |
| 8. HPC6 | Eni S.p.A. / Italy | 0.477 |
| 9. Fugaku | RIKEN Center for Computational Science / Japan | 0.442 |
| 10. Alps | Swiss National Supercomputing Centre / Switzerland | 0.434 |
While China holds the top spot as the world’s fastest supercomputer, Western countries remain dominant too.
Hewlett Packard Enterprise (HPE) remains the dominant system integrator, supplying six of the top ten machines.
On the chip side, AMD powers four of the top ten systems, contributing over 40% of the combined performance of the elite tier.
Real-World Performance and Diverse Benchmarks
While LineShine dominates raw mathematical speed, others still lead in different measurement categories for the world’s fastest supercomputer. The supercomputing community uses multiple tests to look at performance from various angles.
- The HPCG Benchmark: This test measures how systems handle data-intensive, real-world applications rather than pure mathematical equations. LineShine leads here too, taking the number one spot with 22.00 HPCG-Petaflops, beating El Capitan and Japan’s Fugaku.
- The HPL-MxP Benchmark: This measures mixed-precision performance, which is vital for commercial AI workloads. On this test, El Capitan keeps its lead at 16.7 Exaflops due to its dedicated GPU accelerators. LineShine finished fourth here with 7.92 Exaflops, showing the natural limits of a CPU-only architecture when dealing with low-precision AI tasks.
- The Green500 List: Energy efficiency rankings remained completely unchanged. The KAIROS system in France kept its number one spot, achieving 73.28 Gigaflops/Watt using Nvidia Grace Hopper Superchips.
End Note
The crowning of LineShine as the World’s Fastest Supercomputer proves that the global race for computing supremacy is wide open.
By utilizing an independent hardware and software ecosystem, Chinese engineers bypassed international chip sanctions and built a record-breaking exascale machine.
While American technology labs still hold immense power through corporate AI networks, this new entry on the Top500 list ensures that the battle for technological leadership will only intensify.
Maria Isabel Rodrigues
