According to the latest supercomputer ranking from TOP500, El Capitan is the world’s most powerful system, capable of achieving 1,742 petaflops.
In this graphic, Visual Capitalist's Marcus Lu visualizes the performance and power consumption of the world’s top 10 supercomputers.
The data we used to create this graphic is listed in the table below. Figures come from TOP500’s November 2024 ranking.
This ranking highlights America’s position as a global leader in computing power. For instance, the top three supercomputers are located in the U.S., and all of them are classified as exascale systems.
This is a significant change from the 2021 ranking, in which Japan’s Supercomputer Fugaku held the top spot.
The winner of this year’s ranking is El Capitan, which became operational in 2024 and is the third exascale system deployed by the United States.
Sponsored by the U.S. Department of Energy, El Capitan was manufactured by HP Enterprise Cray and features an unknown number of AMD Instict MI300A accelerated processing units (APUs).
An exascale computer is capable of performing at least 1 exaflop (1 quintillion operations per second).
Note that in this ranking, performance is measured in petaflops (1 quadrillion). 1,000 petaflops is equal to 1 exaflop.
In realistic terms, exascale computers can handle massive datasets and solve problems at unprecedented speed. Use cases include:
Since 2021, reports have suggested that China does indeed have its own exascale systems, but is simply withholding information about them from global rankings such as the TOP500.
It’s a well known situation that China has these computers, and they have been operating for a while. They have not run the benchmarks, but [the community has] a general idea of their architectures and capabilities.
- Jack Dongarra, co-founder of TOP500
China’s silence is likely related to its geopolitical strategy. Publicly announcing it has built its own exascale systems could result in further U.S. trade restrictions.
Want to learn more? Check out the Technology category on Voronoi to see more data visualizations like this.