China's LineShine stays No. 1 on TOP500 as new entries hit a record low
The November list keeps the same top 10 as June and adds just 27 new systems, the fewest ever. Tom's Hardware ties the drop to new hardware going to AI data centers.
Key takeaways
- China's LineShine remains No. 1 on the 68th TOP500 list with 2.198 exaflop/s on HPL.
- The top 10 is unchanged since June; El Capitan, Frontier, Aurora and JUPITER Booster follow at No. 2 to No. 5.
- Only 27 systems are new, the fewest ever recorded, TOP500 says.
- Tom's Hardware links the slowdown to new hardware going mainly to AI data centers.
- Korea's Hanggang (No. 12) is the highest new entry; the US Energy Department's Lux debuts at No. 14, Tom's Hardware reports.
China's LineShine supercomputer is still the world's fastest on the TOP500 list, according to the list's 68th edition, released at the SC26 conference in Chicago on Monday. The ten leading systems are unchanged since the June edition, and only 27 systems on the list are new, the lowest number ever recorded, the TOP500 project says.
LineShine, at the National Supercomputing Centre in Shenzhen, reached 2.198 exaflop/s on the High Performance Linpack (HPL) benchmark, according to TOP500. It entered the list in June, ending the run of El Capitan at the top, and is the first system based in China to lead the list since Sunway TaihuLight in 2017. TOP500 says China reported no other large system this time. LineShine also leads the separate HPCG ranking.
How do the top systems compare?
El Capitan at Lawrence Livermore National Laboratory in California stays at No. 2 with 1.809 exaflop/s, followed by Frontier at Oak Ridge National Laboratory with 1.353 exaflop/s and Aurora at Argonne National Laboratory with 1.012 exaflop/s, TOP500 says. Germany's JUPITER Booster at the Jülich Supercomputing Centre is No. 5 and remains the first exascale system in Europe.
Tom's Hardware notes that LineShine uses only CPUs, built on the homegrown LingKun platform with 304-core LX2 processors, while the other four exascale-class machines use accelerators. It puts LineShine's power draw during the benchmark at 42.22 megawatts, against 29.69 megawatts for El Capitan, which Tom's Hardware says makes the US system significantly more power-efficient.
Why are there so few new systems?
The TOP500 project says turnover among the largest systems slowed even as the performance needed to enter the list rose. Tom's Hardware reports that the 27 newcomers make up 5.4% of the list and that the combined HPL performance of all 500 systems grew 4.4% in six months, to 19.57 exaflop/s. The entry point rose 7.3%, to 2.857 petaflop/s, it says.
Tom's Hardware links the low number of new systems to where new computing hardware is going: AI data centers for commercial workloads rather than traditional supercomputers. It cites Epoch AI estimates, quoted by the TOP500 authors, that the AI computing capacity added worldwide each quarter has grown 2.5 times a year.
Which new systems joined?
The highest new entry is Hanggang at South Korea's KISTI Supercomputing Center, at No. 12 with 342.6 petaflop/s, using Nvidia GH200 Grace Hopper Superchips, Tom's Hardware reports. Lux, built by HPE and developed by AMD with the US Department of Energy using AMD Instinct MI355X GPUs, debuts at No. 14 at Oak Ridge, according to a second Tom's Hardware report, which says half of its node-hours each year are reserved for the government's Genesis Mission research effort.
Tom's Hardware also reports that systems powered by AMD now number 202 on the list, up from 177 a year ago. The number of systems with accelerators rose from 275 in June to 291, it says.
Questions readers are asking
What is the world's fastest supercomputer in November 2026?
LineShine at the National Supercomputing Centre in Shenzhen, China, with 2.198 exaflop/s on the HPL benchmark, according to the TOP500 list.
Why did so few new systems join the list?
TOP500 says turnover among large systems slowed. Tom's Hardware links this to most new computing hardware going to commercial AI data centers rather than supercomputers.
How many exascale systems are on the list?
Five: LineShine, El Capitan, Frontier, Aurora and JUPITER Booster, which hold the top five places.
Sources
- #top500
- #supercomputers
- #hpc
- #sc26
- #exascale
Get the day's top AI & tech stories on WhatsApp: Follow the NextFish channel
Get stories like this every morning.
Related
Chips & Infra2 min
Nvidia unveils Rubin HPC, a science-focused GPU due in 2028
Nvidia says the chip will have more than 50,000 CUDA cores but has not given performance, memory or FP64 figures. ServeTheHome says FP64 is the number to watch.
Chips & Infra2 min
US issues rule to stop remote access workarounds to AI chip export limits
The Commerce Department's BIS says the rule targets data centers in third countries. Reports say it adds license rules for five countries, including Singapore and Malaysia.
Chips & Infra2 min
Nvidia board is not considering Huang's return as CEO, CNBC reports
A person familiar with the matter says the board is concentrating on its inquiry into Huang. Tech figures and a petition have called for his return.
Chips & Infra3 min
MLPerf Training v6.1 adds post-training test as Rubin, TPUv7 debut
MLCommons says 24 organizations submitted 250 results. Nvidia claims the fastest times, Lambda reports its own gains, and the first Rubin and Google TPUv7 entries appear.