News

TOP500 Leadership Holds Steady as New System Turnover Falls to a Historic Low

CHICAGO, Illinois - November 16, 2026 - The 68th edition of the TOP500 list of the world's most powerful supercomputers was announced today at SC26 in Chicago. The ten leading systems did not change since the June edition. Just 27 systems are new, the lowest number ever recorded by a wide margin. At the same time, performance thresholds rose across the lower ranks, showing continued growth in installed capability even as turnover among the largest systems slowed.

The Top 10 Is Completely Unchanged

LineShine remains No. 1 with 2.198 Exaflop/s on the High Performance Linpack benchmark. El Capitan follows at No. 2 with 1.809 Exaflop/s, ahead of Frontier at 1.353 Exaflop/s, Aurora at 1.012 Exaflop/s, and JUPITER Booster at 1.000 Exaflop/s. These five systems continue to define the HPL exascale class. HPC7, Eagle, HPC6, Supercomputer Fugaku, and Alps retain positions six through ten, respectively. The unchanged order marks an unusually quiet six months at the top of the ranking after the substantial changes seen in June.

Rank System Site Country HPL (Exaflop/s)
1 LineShine National Supercomputer Center, 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

 

Only 27 Systems Are New

The November list contains only 27 new systems, representing 5.4 percent of the TOP500. This is the lowest level of new-system turnover in the history of the list and is well below the normal movement between semiannual editions. The result does not indicate that performance has stopped growing. Instead, it shows that a larger share of improvement is coming from additions and upgrades below the very top rather than from wholesale replacement of leading installations.

Where the New Hardware Is Going

Considering the marked lack of new HPC systems in the TOP500 while the overall market for advanced computing hardware is booming, one obvious question is where all the new hardware is being used. According to Epoch AI, quarterly additions to global AI computing capacity have grown by roughly 2.5 times per year, while cumulative global AI compute has surpassed the equivalent of 15 million NVIDIA H100 GPUs. As of the fourth quarter of 2025, five major U.S. hyperscalers owned an estimated 71 percent of global AI compute and deployed much of this capacity in mega-clusters often described as AI factories. This suggests that growth has indeed shifted toward AI factories, and from capability systems to capacity systems.

HANGANG Is the Largest Newcomer

HANGANG, installed at the KISTI Supercomputing Center in South Korea, is the highest-ranked new system at No. 12. The HPE Cray EX254n system combines NVIDIA Grace CPUs with GH200 Superchips and delivers 342.6 Petaflop/s on HPL. HANGANG also enters the High Performance Conjugate Gradient ranking at No. 8 with 4.29 HPCG-Petaflop/s, giving it a strong result on a benchmark designed to reflect data-intensive application behavior as well as peak numerical throughput.

Performance Deeper in the List Improved

While the leaders did not move, the performance required to hold a position on the list increased. The No. 500 system now delivers 2.857 Petaflop/s, up 7.3 percent from 2.662 Petaflop/s in June. The threshold at No. 50 rose by 12.0 percent, from 49.85 to 55.81 Petaflop/s, while the No. 100 threshold increased from 21.85 to 22.74 Petaflop/s. The number of systems delivering at least 100 Petaflop/s increased from 25 to 27, and total HPL performance across all 500 systems rose 4.4 percent to 19.57 Exaflop/s. Accelerator-based systems also increased from 275 to 291, or 58.2 percent of the list.

The Green500 Top 10 Is Completely Unchanged

The Green500 energy-efficiency ranking shows the same stability. Its entire Top 10 remains unchanged from June. KAIROS at CALMIP and the University of Toulouse-CNRS in France remains No. 1 at 73.28 Gigaflop/s per Watt, followed by ROMEO-2025 in France at 70.91 Gigaflop/s per Watt and the Levante GPU extension at DKRZ in Germany at 69.43 Gigaflop/s per Watt. An unchanged Green500 Top 10 alongside an unchanged TOP500 Top 10 points to slower turnover in leading system technologies. The current generation continues to deliver incremental gains, but no newly reported architecture has yet displaced the established leaders in either performance or efficiency.

Benchmark Leadership Also Remains Stable

LineShine continues to lead HPCG at 22.00 HPCG-Petaflop/s, followed by El Capitan and Supercomputer Fugaku. On the HPL-MxP mixed-precision benchmark, El Capitan remains first at 16.7 Exaflop/s, ahead of Aurora at 11.6 Exaflop/s and Frontier at 11.4 Exaflop/s. HANGANG's HPCG entry is the principal change near the top of the complementary benchmark rankings.

Taken together, the November results describe a mature phase in the current deployment cycle: leadership is stable, the number of newcomers is exceptionally low, and the most visible progress is occurring deeper in the list. The next shift at the top will therefore depend on when the next generation of large systems and energy-efficient technologies is ready to report measured results.


About the TOP500 List

The TOP500 project began in 1993 as a one-time exercise for a small conference in Mannheim, Germany, followed by a second list compiled later that year for the SC93 conference in the United States. Comparing the two editions revealed how valuable the resulting statistics were, and the project has continued ever since, publishing an updated ranking of the world's most powerful computer systems every June and November.