What is the primary purpose of the TOP500 project?

The TOP500 project serves as a database that ranks and details the 500 most powerful non-distributed computer systems in the world. It aims to provide a reliable basis for tracking and detecting trends in high-performance computing.

What benchmark is used to rank the supercomputers on the TOP500 list?

The rankings are based on the HPL (High-performance LINPACK) benchmark, which is a portable implementation written in Fortran specifically designed for distributed-memory computers.

Who are the key individuals responsible for compiling the TOP500 list?

The TOP500 list is compiled by Jack Dongarra from the University of Tennessee, Erich Strohmaier and Horst Simon from the National Energy Research Scientific Computing Center (NERSC) and Lawrence Berkeley National Laboratory (LBNL), and historically, Hans Meuer of the University of Mannheim until his passing in 2014.

When was the TOP500 project established, and what was its initial goal?

The TOP500 project was established on June 24, 1993. Its goal was to provide meaningful statistics and track trends in high-performance computing by creating a ranked list of the most powerful installed computer systems.

What is the official website for the TOP500 project?

The official website for the TOP500 project is top500.org.

According to the June 2025 TOP500 list, which supercomputer is ranked as the most powerful?

As of June 2025, El Capitan, located in the United States, holds the top position as the most powerful supercomputer on the TOP500 list, achieving 1742 petaFLOPS on the LINPACK benchmarks.

Which country had the most supercomputers on the TOP500 list as of June 2025?

As of June 2025, the United States leads the TOP500 list with the highest number of supercomputers, totaling 175 systems.

What percentage of the total computing power on the TOP500 list did the United States hold as of June 2025?

The United States accounted for 48.4% of the total computing power on the TOP500 list as of June 2025.

What are the Green500 and HPCG benchmarks, and how do they relate to the TOP500?

The Green500 benchmark measures the energy efficiency of supercomputers, while the HPCG (High Performance Conjugate Gradient) benchmark assesses performance on a different type of computational problem, focusing on memory bandwidth and communication. These are additional rankings that provide different perspectives on supercomputer capabilities beyond the primary TOP500 list.

How did the TOP500 project originate in the early 1990s?

The project originated from a need to define and track supercomputing statistics. After initial experiments with metrics like processor count, the University of Mannheim proposed using a detailed listing of installed systems, leading to the collaboration with Jack Dongarra and the use of LINPACK benchmarks.

How has the performance of the top-ranked supercomputer evolved since the TOP500's inception?

Since 1993, the performance of the No. 1 ranked supercomputer has shown consistent and rapid growth, roughly doubling every 14 months, which aligns with the trend predicted by Moore's Law.

What architectural trend has become dominant in recent TOP500 supercomputers?

Heterogeneous computing, which combines different types of processors like CPUs with accelerators such as GPUs (Graphics Processing Units) or Intel's Xeon Phi, has become dominant. This trend is driven by the improved performance-per-watt ratios and higher overall performance these architectures offer.

What is the typical instruction set architecture (ISA) used by most supercomputers on the TOP500 list?

As of recent lists, most supercomputers on the TOP500 are 64-bit systems primarily based on the x86-64 instruction set architecture, with Intel's EMT64 and AMD's AMD64 being the most prevalent CPU types.

What operating system is universally used by the fastest supercomputers listed in the TOP500?

Since the era of the Earth Simulator, operating systems based on Linux have been used by the fastest supercomputers. By November 2017, all supercomputers on the TOP500 list were running an operating system based on the Linux kernel.

When did Windows and Mac OS based systems last feature prominently on the TOP500 list?

Windows-based systems were last significantly present before November 2015, although Ubuntu (a Linux distribution) reappeared on Microsoft Azure systems in 2021. Mac OS computers have not appeared on the list for over a decade.

What was the performance of the Fugaku supercomputer on the HPC-AI benchmark in November 2020?

In November 2020, the ARMv8.2-based Fugaku achieved 2.0 exaflops on the mixed-precision HPC-AI benchmark, marking the first time any system exceeded one exaflop on this metric.

What is the difference between Rmax and Rpeak performance metrics?

Rmax refers to the maximum performance achieved by a supercomputer on the LINPACK benchmark, which is used for ranking. Rpeak represents the theoretical peak performance capability of the system.

Which interconnect technology is most commonly used by the highest-performing supercomputers on the TOP500 list?

InfiniBand is the most utilized interconnect technology based on performance share, accounting for 38% of the systems, while Gigabit Ethernet is more common by the sheer number of computers (54%).

Which countries were in the top three for the number of supercomputers on the TOP500 list as of June 2025?

As of June 2025, the United States led with 175 systems, followed by the European Union (137 systems), and then China (47 systems).

What was the total aggregate computational power (Rmax) of US supercomputers as of June 2025?

The United States held the highest aggregate computational power, reaching 6,696 Petaflops Rmax as of June 2025.

Which supercomputer held the number one spot in the TOP500 in June 2018, and what was its performance?

Summit, an IBM-built system at Oak Ridge National Laboratory, secured the number one position in June 2018 with a performance of 122.3 petaflops/s (PFLOP/s).

What significant shift occurred in HPC manufacturing leadership in June 2018?

In June 2018, Lenovo became the leading HPC manufacturer in terms of installed systems (23.8%), surpassing US-based companies for the first time. This marked a shift in the global landscape of supercomputer production.

What was the objective of the National Strategic Computing Initiative launched in the US?

The National Strategic Computing Initiative, launched via executive order in July 2015, aimed to accelerate the development of exascale computing systems (capable of 1000 petaflops) and to fund research into computing technologies beyond traditional semiconductors.

When did the Frontier supercomputer achieve exascale performance?

The Frontier supercomputer broke the exascale barrier in May 2022, successfully completing more than a quintillion (10^18) 64-bit floating-point calculations per second.

What are common reasons why large supercomputing systems might not appear on the TOP500 list?

Some systems are not submitted to the TOP500 list due to decisions by their operators not to participate, citing concerns about the list's accuracy in reflecting useful work (like Blue Waters), or for security and commercial competitiveness reasons (like OceanLight). Additionally, machines not capable of running or not running the HPL benchmark are excluded.

What is the claimed performance of Google's TPU v5p in Bfloat16 format?

Google's TPU v5p claims a performance exceeding 4 exaflops in Bfloat16 floating-point format, although this is specialized for machine learning workloads and differs from the TOP500's HPL benchmark.

What was the alleged configuration of xAI's 'Colossus' supercomputer?

The xAI 'Colossus' supercomputer allegedly features 100,000 Nvidia H100 GPUs. However, it is reportedly not fully operational due to power supply issues, preventing its inclusion and benchmarking for the TOP500.

Why has China become more secretive about its supercomputing developments in recent years?

Following the US-China trade war, China has increased the secrecy around its new supercomputers and data centers, partly to avoid being targeted by US sanctions, particularly those related to sensitive research areas like nuclear weapons development.

What types of computer architectures have become obsolete or dropped off the TOP500 list?

Over time, architectures such as those using Cell coprocessors (like IBM Roadrunner), Itanium processors, vector processors (like NEC's Earth Simulator), and older systems like Sun Starfire and IBM AIX-based machines have ceased to appear on the TOP500 list as newer technologies have emerged.

What was the last non-Linux operating system to appear on the TOP500 list, and when did it disappear?

The last non-Linux systems were two IBM AIX-based machines running on POWER7 processors, which were ranked in the TOP500 in July 2017 but dropped off the list by November 2017.

What milestone did the June 2022 TOP500 list represent regarding system architecture?

The June 2022 TOP500 list was the first edition where all featured supercomputers were exclusively 64-bit systems.

Which supercomputer was the fastest in Japan as of the November 2023 TOP500 list, and what was its overall rank?

As of November 2023, Fugaku was the fastest supercomputer in Japan, holding the 4th position on the TOP500 list.

What is the role of the LINPACK benchmark in the TOP500 ranking system?

The LINPACK benchmark is a standardized method for measuring the floating-point calculation speed of high-performance computers. Its results directly determine the ranking of systems on the TOP500 list.

How many Nvidia H100 GPUs did Meta AI report using in its AI data centers in March 2024?

In March 2024, Meta AI disclosed the use of 24,576 Nvidia H100 GPUs across two of its AI data centers.

What is the significance of the 'New entry' marker in the TOP500 table?

The 'New entry' marker, often visually indicated, signifies a supercomputer that has recently been added to the TOP500 list or has made a substantial jump in its ranking compared to previous lists.

What information is provided in the 'See also' section of the TOP500 article?

The 'See also' section offers links to related topics and articles, such as computer science, grid computing, Green500, Graph500, and lists of other computer rankings, providing further context and related information.

What does the 'Legend' section of the TOP500 table explain?

The Legend section clarifies the meaning of the various columns used in the TOP500 tables, including Rank, Rmax, Rpeak, Name, Model, Processor, Accelerator, Total Cores, Interconnect, Manufacturer, Site, Country, Year, and Operating System.

What was the performance of Sunway TaihuLight in petaFLOPS/s when it topped the TOP500 list in June 2016?

In June 2016, Sunway TaihuLight achieved 93 petaflops/s (PFLOP/s) on the Linpack benchmark, securing the number one position on the TOP500 list.

What is the difference between Rmax and Rpeak performance?

Rmax represents the actual measured performance of a supercomputer using the LINPACK benchmark, which is the basis for the TOP500 ranking. Rpeak, on the other hand, is the theoretical maximum performance the system is capable of achieving.

Which supercomputer is noted for having the most CPU cores according to the provided text?

The Sunway TaihuLight is identified as the system with the most CPU cores, totaling 10,649,600.

Which supercomputer is noted for having the most GPU/accelerator cores according to the provided text?

The Tianhe-2 is mentioned as having the highest number of GPU/accelerator cores, with 4,554,752.

What is the significance of the 'TOP500' logo image?

The image of the TOP500 logo, typically featuring stylized text, represents the project itself and is used to identify the source of the supercomputer ranking data.

What was the performance of the Cray XK7 'Titan' when it was ranked No. 1 in November 2012?

In November 2012, the Cray XK7 system named Titan, installed at Oak Ridge National Laboratory, achieved 17.59 Petaflop/s on the Linpack benchmark, making it the world's fastest supercomputer at that time.

What is the typical processor architecture for the majority of supercomputers on the TOP500 list?

The majority of supercomputers on the TOP500 list utilize 64-bit processors, predominantly based on the x86-64 instruction set architecture, with Intel's EMT64 and AMD's AMD64 being the most common.

What does the 'Year' column in the TOP500 table typically indicate?

The 'Year' column in the TOP500 table usually signifies the year a supercomputer was installed or underwent its last major upgrade.

What was the performance of the IBM BlueGene/Q 'Sequoia' when it ranked No. 1 in June 2012?

In June 2012, the IBM BlueGene/Q system 'Sequoia' achieved a Linpack benchmark performance of 16.32 petaflop/s, utilizing 1,572,864 cores.

What is the purpose of the 'External links' section in the article?

The 'External links' section provides direct links to the official TOP500 website and specific resources like the LINPACK benchmarks page, allowing readers to access primary data and further information.

Which supercomputer was the fastest in Japan as of the June 2020 TOP500 list, and what was its overall rank?

As of the June 2020 TOP500 list, the Fugaku supercomputer, located in Japan, was ranked No. 1, achieving 442 petaflops on the HPL benchmark.

What is the significance of the 'Sublist Generator' mentioned in the references?

The 'Sublist Generator' is a tool referenced from the TOP500 website that allows users to create custom lists based on various criteria, such as manufacturer, country, or architecture, providing flexible data analysis capabilities.

What trend in processor architecture has been observed for supercomputers dropping off the list?

Older processor architectures like Cell, Itanium, and vector processors have gradually disappeared from the TOP500 list as newer, more powerful, and efficient architectures like x86-64 and ARM have become dominant.

Top500 Wiki: Supercomputing Frontiers: The TOP500 Chronicle

Dive in with Flashcard Learning!

When you are ready...
🎮 Play the Wiki2Web Clarity Challenge Game🎮

TOP500 Overview

The Benchmark

The TOP500 project provides a definitive ranking of the 500 most powerful non-distributed computer systems globally. Published biannually, it relies on the High-Performance LINPACK (HPL) benchmark, a standardized test measuring a system's floating-point calculation capabilities.

Publication Schedule

The list is updated twice a year: in June, coinciding with the International Supercomputing Conference, and in November, presented at the ACM/IEEE Supercomputing Conference. This ensures timely tracking of trends in high-performance computing (HPC).

Current Leader

As of the June 2025 list (the 65th edition), the United States' El Capitan holds the top position. This HPE Cray EX255a system achieved a remarkable 1,742 PetaFLOPS (or 1.742 ExaFLOPS) on the LINPACK benchmark, setting a new standard for computational power.

Evolution of Supercomputing

Genesis of TOP500

Initiated in 1993 by researchers including Jack Dongarra, Horst Simon, and the late Hans Meuer, the TOP500 project emerged from a need for standardized statistics in the rapidly evolving supercomputing landscape. The first list was compiled using available internet data and LINPACK benchmarks.

Performance Growth

Since its inception, the performance of the #1 ranked supercomputer has grown exponentially, roughly doubling every 14 months, a pace often compared to Moore's Law. This relentless advancement reflects continuous innovation in processor technology, interconnects, and parallel processing techniques.

The growth in supercomputing power is staggering. For instance, the fastest system in November 1993, the CM-5/1024, had an Rpeak of 131 GFLOPS. By June 2018, Summit achieved 187 PFLOPS, over 1.4 million times faster than the CM-5, showcasing the dramatic progress in just 25 years.

Key Milestones

The TOP500 list has tracked significant shifts, including the dominance of x86 architectures, the rise of GPUs and accelerators for heterogeneous computing, and the universal adoption of Linux as the operating system for high-performance systems. Innovations like ARM-based processors and specialized interconnects continue to shape the landscape.

Architectural Trends

Processor Dominance

Modern TOP500 systems predominantly utilize 64-bit architectures. Intel's x86-64 CPUs remain prevalent, often paired with AMD's x86-64 processors. Increasingly, ARM-based CPUs are appearing, offering improved energy efficiency and performance-per-watt.

The Rise of Accelerators

Heterogeneous computing, leveraging accelerators like NVIDIA GPUs or Intel's Xeon Phi (though less common now), has become standard. These accelerators significantly boost both raw performance (Rmax) and energy efficiency, driving the push towards Exascale computing.

Operating Systems

Since November 2017, all systems on the TOP500 list have run operating systems based on the Linux kernel. This ubiquity underscores Linux's robustness, flexibility, and suitability for the demanding HPC environment.

Global Supercomputing Landscape

United States Leads

As of June 2025, the United States leads the TOP500 with 175 systems, representing a significant portion of the total computational power. This dominance reflects substantial investment in national research laboratories and advanced computing initiatives.

Distribution by Nation

Following the US, the European Union (as a bloc) and China are major players. Germany, Japan, and France also host a substantial number of high-performance systems, indicating a global distribution of HPC capabilities across leading industrialized nations.

Distribution of supercomputers in the TOP500 list by country (as of June 2025)
Country or territory	Number of systems
United States	175
European Union	137
China	47
Germany	41
Japan	39
France	25
Italy	17
South Korea	15
Canada	13
United Kingdom	13

TOP 10 Supercomputers

The Elite Tier

The top echelon of supercomputing represents the pinnacle of technological achievement. These systems are crucial for scientific research, national security, and complex simulations across various fields.

Top 10 positions of the 65th TOP500 in June 2025
Rank	Rmax Rpeak (PetaFLOPS)	Name	Model	CPU Cores	Accelerator Cores	Total Cores	Manufacturer	Site / Country	Year	OS
1	1,742.00 2,746.38	El Capitan	HPE Cray EX255a	1,051,392	9,988,224 (AMD Instinct MI300A)	11,039,616	HPE	Lawrence Livermore National Laboratory / United States	2024	Linux (TOSS)
2	1,353.00 2,055.72	Frontier	HPE Cray EX235a	614,656	8,451,520 (AMD Instinct MI250X)	9,066,176	HPE	Oak Ridge National Laboratory / United States	2022	Linux (HPE Cray OS)
3	1,012.00 1,980.01	Aurora	HPE Cray EX	1,104,896	8,159,232 (Intel Max 1550)	9,264,128	HPE	Argonne National Laboratory / United States	2023	Linux (SUSE)
4	793.40 930.00	JUPITER	BullSequana XH3000	1,694,592 (Arm Neoverse V2)	3,106,752 (Nvidia Hopper H100)	4,801,344	Atos	EuroHPC JU / Germany	2025	Linux (RHEL)
5	561.20 846.84	Eagle	Microsoft NDv5	172,800	1,900,800 (Nvidia Hopper H100)	2,073,600	Microsoft	Microsoft / United States	2023	Linux (Ubuntu)
6	477.90 606.97	HPC6	HPE Cray EX235a	213,120	2,930,400 (AMD Instinct MI250X)	3,143,520	HPE	Eni S.p.A / Italy	2024	Linux (RHEL)
7	442.01 537.21	Fugaku	Supercomputer Fugaku	7,630,848 (Fujitsu A64FX)	-	7,630,848	Fujitsu	Riken / Japan	2020	Linux (RHEL)
8	434.90 574.84	Alps	HPE Cray EX254n	748,800 (Arm Neoverse V2)	1,372,800 (Nvidia Hopper H100)	2,121,600	HPE	CSCS / Switzerland	2024	Linux (HPE Cray OS)
9	379.70 531.51	LUMI	HPE Cray EX235a	186,624	2,566,080 (AMD Instinct MI250X)	2,752,704	HPE	EuroHPC JU / Finland	2022	Linux (HPE Cray OS)
10	241.20 306.31	Leonardo	BullSequana XH2000	110,592	1,714,176 (Nvidia Ampere A100)	1,824,768	Atos	EuroHPC JU / Italy	2023	Linux (RHEL)

Fastest by Country

National HPC Leaders

This section highlights the leading supercomputer within various countries, showcasing the global distribution of top-tier computational resources and the specific systems driving scientific advancement worldwide.

Fastest supercomputer in TOP500 by country (as of November 2023)
Country/Territory	Fastest Supercomputer (Name)	Rank in TOP500	Rmax / Rpeak (TFlop/s)	Site
United States	El Capitan	1	1,742,000.0 / 2,746,380.0	Lawrence Livermore National Laboratory
Japan	Fugaku	4	442,010.0 / 537,210.0	RIKEN
Finland	LUMI	5	379,700.0 / 531,510.0	Center for Scientific Computing
Italy	Leonardo	6	238,700.0 / 304,470.0	CINECA
Spain	MareNostrum	8	138,200.0 / 265,570.0	Barcelona Supercomputing Center
China	Sunway TaihuLight	11	93,010.0 / 125,440.0	National Supercomputing Center, Wuxi
Netherlands	ISEG	16	46,540.0 / 86,790.0	Nebius
France	Adastra	17	46,100.0 / 61,610.0	GENCI-CINES
Germany	JUWELS (booster module)	18	44,120.0 / 70,980.0	Forschungszentrum Jülich
Saudi Arabia	Shaheen III	20	35,660.0 / 39,610.0	King Abdullah University of Science and Technology

Teacher's Corner

Edit and Print this course in the Wiki2Web Teacher Studio

Edit and Print Materials from this study in the wiki2web studio

Click here to open the "Top500" Wiki2Web Studio curriculum kit

Use the free Wiki2web Studio to generate printable flashcards, worksheets, exams, and export your materials as a web page or an interactive game.

True or False?

Test Your Knowledge!

Gamer's Corner

Are you ready for the Wiki2Web Clarity Challenge?

Learn about top500 while playing the wiki2web Clarity Challenge game.

Unlock the mystery image and prove your knowledge by earning trophies. This simple game is addictively fun and is a great way to learn!

Play now

Explore More Topics

Discover other topics to study!

voter suppression in the united states

dna replication

freyr

world trade organization

dominican republic

vestibule architecture

References

A full list of references for this article are available at the TOP500 Wikipedia page

Feedback & Support

To report an issue with this page, or to find out ways to support the mission, please click here.

Disclaimer

Important Notice

This page was generated by an Artificial Intelligence and is intended for informational and educational purposes only. The content is based on a snapshot of publicly available data from Wikipedia and may not be entirely accurate, complete, or up-to-date.

This is not professional advice. The information provided on this website is not a substitute for professional consultation regarding supercomputing, technology, or scientific research. Always refer to official documentation and consult with qualified experts for specific needs.

The creators of this page are not responsible for any errors or omissions, or for any actions taken based on the information provided herein.

Supercomputing Frontiers

💡 Dive in with Flashcard Learning! 💡

TOP500 Overview 📊

🎯 The Benchmark

⏱️ Publication Schedule

👑 Current Leader

Evolution of Supercomputing ⏳

📜 Genesis of TOP500

📈 Performance Growth

💡 Key Milestones

Architectural Trends 💻

⚙️ Processor Dominance

⚡ The Rise of Accelerators

🐧 Operating Systems

Global Supercomputing Landscape 🌍

🇺🇸 United States Leads

📊 Distribution by Nation

TOP 10 Supercomputers 🏆

🥇 The Elite Tier

Fastest by Country 📍

🗺️ National HPC Leaders

Teacher's Corner 🧑‍🏫

Edit and Print this course in the Wiki2Web Teacher Studio

True or False? 🤔

❓ Test Your Knowledge! ❓

Gamer's Corner 🎮

Are you ready for the Wiki2Web Clarity Challenge?

Explore More Topics

📜 Discover other topics to study!

References

📜 References

Feedback & Support 👍

To report an issue with this page, or to find out ways to support the mission, please click here.

Disclaimer ⚠️

📜 Important Notice

Dive in with Flashcard Learning!

TOP500 Overview

The Benchmark

Publication Schedule

Current Leader

Evolution of Supercomputing

Genesis of TOP500

Performance Growth

Key Milestones

Architectural Trends

Processor Dominance

The Rise of Accelerators

Operating Systems

Global Supercomputing Landscape

United States Leads

Distribution by Nation

TOP 10 Supercomputers

The Elite Tier

Fastest by Country

National HPC Leaders

Teacher's Corner

True or False?

Test Your Knowledge!

Gamer's Corner

Discover other topics to study!

References

Feedback & Support

Disclaimer

Important Notice