TOP500
Key people	Erich Strohmaier Jack Dongarra Horst D. Simon Hans Meuer
Established	24 June 1993; 32 years ago (1993-06-24)
Website	top500.org

TheTOP500 project ranks and details the 500 most powerful non-distributed computer systems in the world. The project was started in 1993 and publishes an updated list of thesupercomputers twice a year. The first of these updates always coincides with theInternational Supercomputing Conference in June, and the second is presented at theACM/IEEE Supercomputing Conference in November. The project aims to provide a reliable basis for tracking and detecting trends in high-performance computing and bases rankings onHPL benchmarks,^[1] a portable implementation of the high-performanceLINPACK benchmark written inFortran fordistributed-memory computers.

The most recent edition of TOP500 was published in June 2025 as the 65th edition of TOP500, while the next edition of TOP500 will be published in November 2025 as the 66th edition of TOP500. As of June 2025, the United States'El Capitan is the most powerful supercomputer in the TOP500, reaching 1742petaFlops (1.742 exaFlops) on the LINPACK benchmarks.^[2] As of submitted data until June 2025, the United States has the highest number of systems with 175 supercomputers; China is in second place with 47, and Germany is third at 41; the United States has by far the highest share of total computing power on the list (48.4%).^[3] Due to secrecy of the latest Chinese programs, publicly known supercomputer performance share in China represents only 2% that of global as of June 2025.^[3][4][5]

The TOP500 list is compiled byJack Dongarra of theUniversity of Tennessee,Knoxville, Erich Strohmaier and Horst Simon of theNational Energy Research Scientific Computing Center (NERSC) andLawrence Berkeley National Laboratory (LBNL), and, until his death in 2014,Hans Meuer of theUniversity of Mannheim,Germany.^{[citation needed]} The TOP500 project also includes lists such asGreen500 (measuring energy efficiency) andHPCG (measuring I/O bandwidth).^[6]

In the early 1990s, a new definition ofsupercomputer was needed to produce meaningful statistics. After experimenting with metrics based on processor count in 1992, the idea arose at theUniversity of Mannheim to use a detailed listing of installed systems as the basis. In early 1993,Jack Dongarra was persuaded to join the project with hisLINPACK benchmarks. A first test version was produced in May 1993, partly based on data available on the Internet, including the following sources:^[7][8]

• "List of the World's Most Powerful Computing Sites" maintained by Gunter Ahrendt^[9]
• David Kahaner, the director of the Asian Technology Information Program (ATIP);^[10] published a report in 1992, titled "Kahaner Report on Supercomputer in Japan"^[8] which had an immense amount of data.^{[citation needed]}

The information from those sources was used for the first two lists. Since June 1993, the TOP500 is produced bi-annually based on site and vendor submissions only. Since 1993, performance of the No. 1 ranked position has grown steadily in accordance withMoore's law, doubling roughly every 14 months. In June 2018,Summit was fastest with an Rpeak^[11] of 187.6593 PFLOPS. For comparison, this is over 1,432,513 times faster than theConnection Machine CM-5/1024 (1,024 cores), which was the fastest system in November 1993 (twenty-five years prior) with an Rpeak of 131.0 GFLOPS.^[12]

While Intel, or at least thex86-64 CPU architecture has previously dominated the supercomputer list, by now AMD has more systems using that same architecture on top10, including 1st and 2nd place. AndMicrosoft Azure has 8 systems on top100, thereof only two with Intel CPUs, including though its most performant by far in 4th place (previously 3rd place). AMDs CPUs are usually coupled with AMD's GPU accelerators, while Intel's CPUs have historically been very often coupled with NVidia's GPU, though current Intel's third place (previously 2nd place) system notably uses Intel Data Center GPU Max. Arm-based system are also notable on the list in 4th, 7th (Fugaku, previously nr. 1) and 8th place and in total at least 23 not just from Fujitsu that introduced Arm-based the top spot; Nvidia has others with their "Superchip" CPU, not just GPUs.

As of June 2022^[update], all supercomputers on TOP500 are64-bit supercomputers, mostly based onCPUs with thex86-64instruction set architecture, 384 of which areIntelEMT64-based and 101 of which areAMDAMD64-based, with the latter including the top eight supercomputers. 15 other supercomputers are all based onRISC architectures, including six based onARM64 and seven based on thePower ISA used byIBM Power microprocessors.^{[citation needed]}

In recent years,heterogeneous computing has dominated the TOP500, mostly usingNvidia'sgraphics processing units (GPUs) or Intel's x86-basedXeon Phi ascoprocessors. This is because of betterperformance per watt ratios and higher absolute performance. AMD GPUs have taken the top 1 and displaced Nvidia in top 10 part of the list. The recent exceptions include the aforementionedFugaku,Sunway TaihuLight, andK computer.Tianhe-2A is also an interesting exception, as USsanctions prevented use of Xeon Phi; instead, it was upgraded to use the Chinese-designedMatrix-2000^[13] accelerators.^{[citation needed]}

Two computers which first appeared on the list in 2018 were based on architectures new to the TOP500. One was a new x86-64microarchitecture from Chinese manufacturer Sugon, usingHygon Dhyana CPUs (these resulted from a collaboration with AMD, and are a minor variant ofZen-basedAMD EPYC) and was ranked 38th, now 117th,^[14] and the other was the firstARM-based computer on the list – usingCavium ThunderX2 CPUs.^[15] Before the ascendancy of32-bitx86 and later64-bitx86-64 in the early 2000s, a variety of RISC processor families made up most TOP500 supercomputers, includingSPARC,MIPS,PA-RISC, andAlpha.

All the fastest supercomputers since theEarth Simulator supercomputer (gained top spot in 2002, kept it for 2 and a half years until June 2004, was decommissioned in 2009; though other non-Linux on the list for longer) have used operating systems based onLinux. Since November 2017^[update], all the listed supercomputers use an operating system based on theLinux kernel.^[16][17]

Since November 2015, no computer on the list runsWindows (while Microsoft reappeared on the list in 2021 withUbuntu based on Linux). In November 2014,Windows Azure^[18] cloud computer was no longer on the list of fastest supercomputers (its best rank was 165th in 2012), leaving theShanghai Supercomputer Center'sMagic Cube as the only Windows-based supercomputer on the list, until it also dropped off the list. It was ranked 436th in its last appearance on the list released in June 2015, while its best rank was 11th in 2008.^[19] There are no longer anyMac OS computers on the list. It had at most five such systems at a time, one more than the Windows systems that came later, while the total performance share for Windows was higher. Their relative performance share of the whole list was however similar, and never high for either. In 2004, theSystem X supercomputer based onMac OS X (Xserve, with 2,200PowerPC 970 processors) once ranked 7th place.^[20]

It has been well over a decade sinceMIPS systems dropped entirely off the list^[21] though theGyoukou supercomputer that jumped to 4th place^[22] in November 2017 had a MIPS-based design as a small part of the coprocessors. Use of 2,048-core coprocessors (plus 8× 6-core MIPS, for each, that "no longer require to rely on an external Intel Xeon E5 host processor"^[23]) made the supercomputer much more energy efficient than the other top 10 (i.e. it was 5th onGreen500 and other suchZettaScaler-2.2-based systems take first three spots).^[24] At 19.86 million cores, it was by far the largest system by core-count, with almost double that of the then-bestmanycore system, the ChineseSunway TaihuLight.

As of November 2025^[update], the number one supercomputer isEl Capitan, the leader on Green500 is KAIROS, a Bull Sequana XH3000 system using the Nvidia Grace Hopper GH200 Superchip. EuroHPC's JUPITER became Europe's first system to reach the exascale milestone. In June 2022, the top 4 systems ofGraph500 used both AMD CPUs and AMD accelerators. After an upgrade, for the 56th TOP500 in November 2020,

Fugaku grew its HPL performance to 442 petaflops, a modest increase from the 416 petaflops the system achieved when it debuted in June 2020. More significantly, theARMv8.2 based Fugaku increased its performance on the new mixed precision HPC-AI benchmark to 2.0 exaflops, besting its 1.4 exaflops mark recorded six months ago. These represent the first benchmark measurements above one exaflop for any precision on any type of hardware.^[25]

Summit, a previously fastest supercomputer, is currently highest-ranked IBM-made supercomputer; with IBMPOWER9 CPUs.Sequoia became the last IBMBlue Gene/Q model to drop completely off the list; it had been ranked 10th on the 52nd list (and 1st on the June 2012, 41st list, after an upgrade).

For the first time, all 500 systems deliver a petaflop or more on the High Performance Linpack (HPL) benchmark, with the entry level to the list now at 1.022 petaflops." However, for a different benchmark "Summit and Sierra remain the only two systems to exceed a petaflop on theHPCG benchmark, delivering 2.9 petaflops and 1.8 petaflops, respectively. The average HPCG result on the current list is 213.3 teraflops, a marginal increase from 211.2 six months ago.^[26]

Microsoft is back on the TOP500 list with sixMicrosoft Azure instances (that use/are benchmarked withUbuntu, so all the supercomputers are still Linux-based), with CPUs and GPUs from same vendors, the fastest one currently 11th,^[27] and another older/slower previously made 10th.^[28] And Amazon with one AWS instance currently ranked 64th (it was previously ranked 40th). The number of Arm-based supercomputers is 6; currently all Arm-based supercomputers use the same Fujitsu CPU as in the number 2 system, with the next one previously ranked 13th, now 25th.^[29]

Top 10 positions of the 66th TOP500 in November 2025^[30]

Rank (previous)	Rmax Rpeak (PetaFLOPS)	Name	Model	CPU cores	Accelerator (e.g. GPU) cores	Total cores (CPUs + accelerators)	Interconnect	Manufacturer	Site country	Year	Operating system
1	1,809.00 2,821.10	El Capitan	HPE Cray EX255a	1,080,000 (45,000 × 24-coreOptimized 4th Generation EPYC 24C @1.8 GHz)	10,260,000 (45,000 × 228 AMDInstinct MI300A)	11,340,000	Slingshot-11	HPE	Lawrence Livermore National Laboratory  United States	2024	Linux (TOSS)
2	1,353.00 2,055.72	Frontier	HPE Cray EX235a	614,656 (9,604 × 64-coreOptimized 3rd Generation EPYC 64C @2.0 GHz)	8,451,520 (38,416 × 220 AMDInstinct MI250X)	9,066,176	Slingshot-11	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 (21,248 × 52-core IntelXeon Max 9470 @2.4 GHz)	8,159,232 (63,744 × 128 IntelMax 1550)	9,264,128	Slingshot-11	HPE	Argonne National Laboratory  United States	2023	Linux (SUSE Linux Enterprise Server 15 SP4)
4	1,000.00 1,226.28	JUPITER	BullSequana XH3000	1,694,592 (23,536 × 72-Arm Neoverse V2 coresNvidia Grace @3 GHz)	3,106,752 (23,536 × 132 NvidiaHopper H100)	4,801,344	Quad-rail NVIDIA NDR200Infiniband	Atos	EuroHPC JU  European Union,Jülich, Germany	2025	Linux (RHEL)
5	561.20 846.84	Eagle	Microsoft NDv5	172,800 (3,600 × 48-core IntelXeon Platinum 8480C @2.0 GHz)	1,900,800 (14,400 × 132 NvidiaHopper H100)	2,073,600	NVIDIA Infiniband NDR	Microsoft	Microsoft  United States	2023	Linux (Ubuntu 22.04 LTS)
6	477.90 606.97	HPC6	HPE Cray EX235a	213,120 (3,330 × 64-coreOptimized 3rd Generation EPYC 64C @2.0 GHz)	2,930,400 (13,320 × 220 AMDInstinct MI250X)	3,143,520	Slingshot-11	HPE	Eni S.p.A  European Union,Ferrera Erbognone, Italy	2024	Linux (RHEL 8.9)
7	442.01 537.21	Fugaku	Supercomputer Fugaku	7,630,848 (158,976 × 48-coreFujitsu A64FX @2.2 GHz)	-	7,630,848	Tofu interconnect D	Fujitsu	Riken Center for Computational Science  Japan	2020	Linux (RHEL)
8	434.90 574.84	Alps	HPE Cray EX254n	748,800 (10,400 × 72-Arm Neoverse V2 coresNvidia Grace @3.1 GHz)	1,372,800 (10,400 × 132 NvidiaHopper H100)	2,121,600	Slingshot-11	HPE	CSCS Swiss National Supercomputing Centre  Switzerland	2024	Linux (HPE Cray OS)
9	379.70 531.51	LUMI	HPE Cray EX235a	186,624 (2,916 × 64-coreOptimized 3rd Generation EPYC 64C @2.0 GHz)	2,566,080 (11,664 × 220 AMDInstinct MI250X)	2,752,704	Slingshot-11	HPE	EuroHPC JU  European Union,Kajaani, Finland	2022	Linux (HPE Cray OS)
10	241.20 306.31	Leonardo	BullSequana XH2000	110,592 (3,456 × 32-coreXeon Platinum 8358 @2.6 GHz)	1,714,176 (15,872 × 108Nvidia Ampere A100)	1,824,768	Quad-rail NVIDIA HDR100Infiniband	Atos	EuroHPC JU  European Union,Bologna, Italy	2023	Linux (RHEL 8)[31]

Legend:^[32]

• Rank – Position within the TOP500 ranking. In the TOP500 list table, the computers are ordered first by their Rmax value. In the case of equal performances (Rmax value) for different computers, the order is by Rpeak. For sites that have the same computer, the order is by memory size and then alphabetically.
• Rmax – The highest score measured using theLINPACK benchmarks suite. This is the number that is used to rank the computers. Measured inquadrillions of 64-bitfloating point operations persecond, i.e., petaFLOPS.^[33]
• Rpeak – This is the theoretical peak performance of the system. Computed in petaFLOPS.
• Name – Some supercomputers are unique, at least on its location, and are thus named by their owner.
• Model – The computing platform as it is marketed.
• Processor – Theinstruction set architecture or processormicroarchitecture, alongside GPU and accelerators when available.
• Interconnect – Theinterconnect between computing nodes.InfiniBand is most used (38%) by performance share, whileGigabit Ethernet is most used (54%) by number of computers.
• Manufacturer – The manufacturer of the platform and hardware.
• Site – The name of the facility operating the supercomputer.
• Country – The country in which the computer is located.
• Year – The year of installation or last major update.
• Operating system – The operating system that the computer uses.

Numbers below represent the number of computers in the TOP500 that are in each of the listed countries or territories. As of 2025, United States has the most supercomputers on the list, with 171 machines. The United States has the highest aggregate computational power at 6,626 Petaflops Rmax with Japan second (1,283 Pflop/s) and Germany third (1,129 Pflop/s).

Distribution of supercomputers in the TOP500 list by country (as of November 2025)^[3]

Country or territory	Number of systems
United States	171
European Union	126
Japan	43
Germany	40
China	40
France	22
Canada	19
Italy	18
South Korea	15
Taiwan	10
Brazil	10
United Kingdom	9
Norway	9
Sweden	8
Poland	8
Netherlands	7
Saudi Arabia	7
India	6
United Arab Emirates	5
Russia	5
Singapore	5
Australia	4
Switzerland	3
Finland	3
Spain	3
Israel	3
Czechia	3
Austria	2
Kazakhstan	2
Slovakia	2
Thailand	2
Turkey	2
Slovenia	2
Ireland	2
Denmark	1
Vietnam	1
Iceland	1
Luxembourg	1
Argentina	1
Bulgaria	1
Portugal	1
Morocco	1
Hungary	1
Belgium	1

Distribution of supercomputers in the TOP500 list by country and by year

Country / Region	Nov 2025[3]	Jun 2025[3]	Nov 2024[3]	Jun 2024[3]	Nov 2023[3]	Jun 2023[3]	Nov 2022[3]	Jun 2022[3]	Nov 2021[3]	Jun 2021[3]	Nov 2020[3]	Jun 2020[3]	Nov 2019[3]	Jun 2019[3]	Nov 2018[3]	Jun 2018[3]	Nov 2017[3]	Jun 2017[3]	Nov 2016[3]	Jun 2016[3]	Nov 2015[3]	Jun 2015[3]	Nov 2014[3]	Jun 2014[3]	Nov 2013[3]	Jun 2013[3]	Nov 2012[3]	Jun 2012[3]	Nov 2011[3]	Jun 2011[3]	Nov 2010[3]	Jun 2010[3]	Nov 2009[3]	Jun 2009[3]	Nov 2008[3]	Jun 2008[3]	Nov 2007[3]	Jun 2007[3]	Nov 2006[3]
United States	171	175	173	171	161	150	127	128	149	122	113	114	117	116	109	124	143	168	171	165	199	233	231	232	264	252	251	252	263	255	274	282	277	291	290	257	283	281	309
EU	126	128	129	123	112	103	101	92	83	93	79	79	87	92	91	93	86	99	95	93	94	122	110	103	89	97	89	96	95	109	108	126	137	134	140	169	133	115	82
Japan	43	39	34	29	32	33	31	33	32	34	34	29	29	28	31	36	35	33	27	29	37	40	32	30	28	30	32	35	30	26	26	18	16	15	17	22	20	23	30
Germany	40	41	40	40	36	36	34	31	26	23	17	16	16	13	17	21	21	28	31	26	33	37	26	22	20	19	19	20	20	30	26	24	27	29	25	46	31	24	18
China	40	47	63	80	104	134	162	173	173	188	214	226	228	220	227	206	202	160	171	168	109	37	61	76	63	66	72	68	74	61	41	24	21	21	15	12	10	13	18
France	22	25	24	24	23	24	24	22	19	16	18	19	18	20	18	18	18	18	20	18	18	27	30	27	22	23	21	22	23	25	26	27	26	23	26	34	17	13	12
Canada	19	13	9	10	10	10	10	14	11	11	12	12	9	8	9	6	5	6	1	1	6	6	6	9	10	9	11	10	9	8	6	7	9	8	2	2	5	10	8
Italy	18	17	14	11	12	7	7	6	6	6	6	7	5	5	6	5	6	8	6	5	4	4	3	5	5	6	7	8	4	5	6	7	6	6	11	6	6	5	8
South Korea	15	15	13	13	12	8	8	6	7	5	3	3	3	5	6	7	5	8	4	7	10	9	9	8	5	4	4	3	3	4	3	1	2	0	1	1	1	5	6
Taiwan	10	8	7	6	5	2	2	2	2	2	3	2	2	2	2	1	1	0	0	0	0	1	1	1	1	1	3	3	2	2	0	0	0	1	2	3	11	10	2
Brazil	10	9	9	8	9	9	8	6	5	6	4	4	3	3	1	1	0	2	3	4	6	6	4	4	3	3	2	3	2	2	2	1	1	0	2	1	1	2	4
United Kingdom	9	13	14	16	15	14	15	12	11	11	12	10	11	18	20	22	15	17	13	11	18	29	30	30	23	29	24	25	27	27	25	38	45	44	46	53	48	42	30
Norway	9	9	6	5	5	4	3	2	1	3	3	3	2	0	1	1	1	1	1	1	1	2	3	3	3	3	3	3	0	1	3	2	2	2	2	2	3	2	3
Sweden	8	9	8	7	6	6	6	5	4	3	2	2	2	2	4	3	5	5	4	5	3	5	5	3	5	7	6	4	3	5	6	8	7	10	8	9	7	10	1
Poland	8	7	8	8	4	3	3	5	4	4	2	1	1	1	4	4	5	6	7	6	6	7	2	2	2	3	4	5	6	5	6	5	3	4	6	3	1	0	0
Netherlands	7	7	10	9	10	8	8	6	11	16	15	15	15	13	6	9	6	4	3	3	2	3	5	5	3	2	0	0	0	1	2	4	3	3	3	5	6	8	2
Saudi Arabia	7	6	7	8	7	6	6	6	6	6	5	3	3	3	3	4	4	6	5	5	6	7	4	4	3	4	3	3	3	4	6	4	4	2	0	0	0	2	4
India	6	6	6	4	4	4	3	3	3	3	3	2	2	3	4	5	4	4	5	9	11	11	9	9	12	11	8	5	2	2	4	5	3	6	8	6	9	8	10
United Arab Emirates	5	4	3	2	1	1	2	2	2	2	2	2	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	1	1
Russia	5	6	6	7	7	7	7	7	7	3	2	2	3	2	3	4	3	3	5	7	7	8	9	5	5	8	8	5	5	12	11	11	8	5	8	9	7	5	2
Singapore	5	5	4	3	3	3	3	3	1	4	4	4	4	5	3	2	1	1	1	1	0	0	0	0	0	0	0	1	1	2	2	1	1	1	0	0	1	2	2
Australia	4	4	4	5	6	5	5	5	3	2	2	2	3	5	5	5	4	4	3	5	4	6	9	6	5	5	7	6	4	6	4	1	1	1	1	1	1	4	4
Switzerland	3	4	5	5	3	4	4	4	3	3	3	2	2	4	2	3	3	3	4	3	6	6	7	6	5	4	4	1	3	4	4	5	5	4	4	6	7	5	5
Finland	3	3	3	3	3	3	3	4	3	2	2	2	1	2	1	1	2	3	2	5	2	2	3	2	2	2	3	1	1	2	1	3	2	1	1	1	5	3	1
Spain	3	3	3	3	3	1	1	1	1	1	1	1	2	2	2	2	1	1	1	1	2	2	2	2	2	3	2	4	3	2	3	3	6	5	6	7	9	6	7
Israel	3	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	2	2	2	2	1	3	3	2	0	2	2	1	1	0	0	0	2
Czechia	3	3	3	3	2	2	2	2	2	2	1	1	1	1	1	1	1	1	1	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Austria	2	2	3	2	2	2	2	2	1	1	1	1	1	1	0	0	2	3	3	5	1	1	1	1	1	1	1	1	2	2	1	2	8	5	0	0	0	0	0
Kazakhstan	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Slovakia	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Thailand	2	2	2	1	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Turkey	2	2	2	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	2	1	1
Slovenia	2	2	2	2	2	2	2	2	2	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	1	1	1	1	0	0	0
Ireland	2	2	4	4	4	5	5	3	1	14	14	14	14	13	12	7	4	2	1	3	0	0	0	1	2	0	0	3	3	1	1	1	1	1	1	1	0	0	1
Denmark	1	1	1	0	0	0	0	0	0	0	0	0	0	1	0	0	0	2	2	2	2	2	2	1	1	1	1	1	2	2	2	3	0	0	3	0	1	0	1
Vietnam	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0
Iceland	1	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Luxembourg	1	1	1	2	2	2	2	2	2	2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0
Argentina	1	1	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Bulgaria	1	2	2	2	2	1	1	1	1	1	0	0	0	0	0	0	0	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	1	1	1	0	0	0	0
Portugal	1	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Morocco	1	1	1	1	1	1	1	1	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Hungary	1	1	1	1	1	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	1	1	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Belgium	1	1	1	1	1	1	1	0	0	0	0	0	0	0	0	0	0	1	1	2	0	1	1	2	1	1	1	2	1	2	2	0	1	1	2	2	1	4	1
Hong Kong	0	0	0	0	0	0	0	0	0	1	1	1	1	1	0	0	0	0	0	0	1	1	1	1	2	1	0	0	0	0	1	1	1	1	0	0	0	0	0
South Africa	0	0	0	0	0	0	0	0	0	0	0	0	0	3	2	1	1	1	1	1	0	0	0	0	0	0	0	1	1	0	0	1	1	0	1	0	0	1	2
New Zealand	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	1	1	3	1	1	0	0	0	0	0	0	0	0	0	5	7	8	5	4	6	1	1	1
Mexico	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	1	1	0	0	0	0	1	0	0	0	0	0	0	1	1	0	0	2	1
Croatia	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Greece	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
Malaysia	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	1	0	0	0	0	0	0	0	0	1	1	1	2	3	4	3
Slovak Republic	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0	0	0	0	0	0	0	0	0	0
Cyprus	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0	0	0
Egypt	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	0	0
Indonesia	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	1	0
Philippines	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0

(As of November 2025^[34])

Country/Territory	Fastest supercomputer of country/territory (name)	Rank in TOP500	Rmax Rpeak (PFlop/s)	Site
United States	El Capitan	1	1,809.00 2,821.10	Lawrence Livermore National Laboratory
Japan	Fugaku	7	442.01 537.21	RIKEN
Finland	LUMI	9	379.70 531.51	Center for Scientific Computing
Italy	HPC6	6	477.90 606.97	Eni S.p.A.
Spain	MareNostrum	14	175.30 249.44	Barcelona Supercomputing Center
China	Sunway TaihuLight	24	93.01 125.44	National Supercomputing Center, Wuxi
Netherlands	ISEG-2	13	202.40 338.49	Nebius
France	CEA-HE	26	90.79 171.26	CEA
Germany	Jupiter Booster	4	1,000.00 1,226.28	Forschungszentrum Jülich
Saudi Arabia	Shaheen III	18	122.80 155.21	King Abdullah University of Science and Technology
South Korea	SSC-24	21	106.20 151.10	Samsung Electronics
Australia	Setonix	66	27.16 35.00	Pawsey Supercomputing Centre
Sweden	DeepL Mercury	82	21.85 33.85	DeepL SE
Russia	Chervonenkis	83	21.53 29.42	Yandex
Switzerland	Alps	8	434.90 574.84	Swiss National Supercomputing Centre
United Kingdom	Isambard-AI phase 2	11	216.50 278.58	University of Bristol
Brazil	Harpia	36	56.60 120.38	Petróleo Brasileiro S.A
Taiwan	Nano 4	29	81.55 117.92	National Center for High-Performance Computing
Luxembourg	MeluXina - Accelerator Module	158	10.52 15.29	LuxProvide
India	PowerEdge XE9680	28	84.31 102.82	Shakti Cloud, Yotta Data Services Private Limited
Thailand	THE CRUST 2.5	127	13.85 21.68	PTT Exploration and Production
Canada	Sovereign AI Factory	78	22.74 27.44	Telus Communications
UAE	SuperPOD	37	55.81 87.27	Group 42
Czechia	C24	193	8.40 8.87	Škoda Auto
Poland	Helios GPU	96	19.14 30.44	Cyfronet
Norway	Olivia (GPU)	134	13.20 16.80	UNINETT Sigma2 AS
Bulgaria	Discoverer	291	4.52 5.94	Consortium Petascale Supercomputer Bulgaria
Argentina	Clementina XXI	266	5.39 12.58	Servicio Meteorológico Nacional
Slovenia	VEGA HPC CPU	337	3.82 5.37	IZUM
Ireland	AIC1	357	3.55 6.97	Software Company MIR
Singapore	Hopper NUS	120	15.05 24.99	National University of Singapore
Morocco	Toubkal	385	3.16 5.01	Mohammed VI Polytechnic University - African Supercomputing Centre
Hungary	Komondor	410	3.10 4.51	Governmental Information Technology Development Agency (KIFÜ)
Austria	MUSICA Phase 1	60	31.84 53.76	Austrian Scientific Computing (ASC)
Belgium	Lucia	459	2.78 3.97	Cenaero

This sectionneeds additional citations forverification. Please helpimprove this article byadding citations to reliable sources in this section. Unsourced material may be challenged and removed. Find sources: "TOP500" – news ·newspapers ·books ·scholar ·JSTOR(September 2019) (Learn how and when to remove this message)

• HPE CrayEl Capitan (Lawrence Livermore National Laboratory United States, November 2024 – Present)^[35]
• HPE CrayFrontier (Oak Ridge National Laboratory United States, June 2022 – November 2024)^[36]
• SupercomputerFugaku (Riken Center for Computational Science Japan, June 2020 – June 2022)^[37]
• IBMSummit (Oak Ridge National Laboratory United States, June 2018 – June 2020)^[38]
• NRCPCSunway TaihuLight (National Supercomputing Center in Wuxi China, June 2016 – November 2017)
• NUDTTianhe-2A (National Supercomputing Center of Guangzhou China, June 2013 – June 2016)
• CrayTitan (Oak Ridge National Laboratory United States, November 2012 – June 2013)^[39]
• IBMSequoiaBlue Gene/Q (Lawrence Livermore National Laboratory United States, June 2012 – November 2012)^[40]
• FujitsuK computer (Riken Advanced Institute for Computational Science Japan, June 2011 – June 2012)
• NUDTTianhe-1A (National Supercomputing Center of Tianjin China, November 2010 – June 2011)
• CrayJaguar (Oak Ridge National Laboratory United States, November 2009 – November 2010)
• IBMRoadrunner (Los Alamos National Laboratory United States, June 2008 – November 2009)
• IBMBlue Gene/L (Lawrence Livermore National Laboratory United States, November 2004 – June 2008)^[41]
• NECEarth Simulator (Earth Simulator Center Japan, June 2002 – November 2004)
• IBMASCI White (Lawrence Livermore National Laboratory United States, November 2000 – June 2002)
• IntelASCI Red (Sandia National Laboratories United States, June 1997 – November 2000)^[42]
• HitachiCP-PACS (University of Tsukuba Japan, November 1996 – June 1997)
• HitachiSR2201 (University of Tokyo Japan, June 1996 – November 1996)
• FujitsuNumerical Wind Tunnel (National Aerospace Laboratory of Japan Japan, November 1994 – June 1996)
• IntelParagon XP/S140 (Sandia National Laboratories United States, June 1994 – November 1994)
• FujitsuNumerical Wind Tunnel (National Aerospace Laboratory of Japan Japan, November 1993 – June 1994)
• TMCCM-5 (Los Alamos National Laboratory United States, June 1993 – November 1993)

By number of systems as of November 2025^[update]:^[43]

Note: All operating systems of the TOP500 systems areLinux-family based, but Linux above is generic Linux.

El Capitan is the system with the most CPU cores (11,340,000).El Capitan has the most GPU/accelerator cores (10,260,000).Aurora is the system with the greatest power consumption with 38,698 kilowatts.

In November 2014, it was announced that the United States was developing two new supercomputers to exceed China's Tianhe-2 in its place as world's fastest supercomputer. The two computers,Sierra andSummit, will each exceed Tianhe-2's 55 peak petaflops. Summit, the more powerful of the two, will deliver 150–300 peak petaflops.^[44] On 10 April 2015, US government agencies banned selling chips, from Nvidia to supercomputing centers in China as "acting contrary to thenational security ... interests of the United States";^[45] and Intel Corporation from providing Xeon chips to China due to their use, according to the US, in researching nuclear weapons – research to which USexport control law bans US companies from contributing – "The Department of Commerce refused, saying it was concerned about nuclear research being done with the machine."^[46]

On 29 July 2015,President Obama signed an executive order creating aNational Strategic Computing Initiative calling for the accelerated development of anexascale (1000 petaflop) system and funding research into post-semiconductor computing.^[47]

In June 2016, Japanese firm Fujitsu announced at theInternational Supercomputing Conference that its futureexascale supercomputer will feature processors of its own design that implement theARMv8 architecture. The Flagship2020 program, by Fujitsu for RIKEN plans to break the exaflops barrier by 2020 through theFugaku supercomputer, (and "it looks like China and France have a chance to do so and that the United States is content – for the moment at least – to wait until 2023 to break through the exaflops barrier."^[48]) These processors will also implement extensions to the ARMv8 architecture equivalent to HPC-ACE2 that Fujitsu is developing withArm.^[48]

In June 2016, Sunway TaihuLight became the No. 1 system with 93 petaflop/s (PFLOP/s) on the Linpack benchmark.^[49]

In November 2016, Piz Daint was upgraded, moving it from 8th to 3rd, leaving the US with no systems under the TOP3 for the 2nd time.^[50][51]

Inspur, based out ofJinan, China, is one of the largest HPC system manufacturers. As of May 2017^[update],Inspur has become the third manufacturer to have manufactured a 64-way system – a record that has previously been held byIBM andHP. The company has registered over $10B in revenue and has provided a number of systems to countries such as Sudan, Zimbabwe, Saudi Arabia and Venezuela. Inspur was also a major technology partner behind both theTianhe-2 andTaihu supercomputers, occupying the top 2 positions of the TOP500 list up until November 2017. Inspur andSupermicro released a few platforms aimed at HPC using GPU such as SR-AI and AGX-2 in May 2017.^[52]

In June 2018, Summit, an IBM-built system at the Oak Ridge National Laboratory (ORNL) in Tennessee, US, took the No. 1 spot with a performance of 122.3 petaflop/s (PFLOP/s), and Sierra, a very similar system at the Lawrence Livermore National Laboratory, CA, US took #3. These systems also took the first two spots on the HPCG benchmark. Due to Summit and Sierra, the US took back the lead as consumer of HPC performance with 38.2% of the overall installed performance while China was second with 29.1% of the overall installed performance. For the first time ever, the leading HPC manufacturer was not a US company. Lenovo took the lead with 23.8% of systems installed. It is followed by HPE with 15.8%, Inspur with 13.6%, Cray with 11.2%, and Sugon with 11%.^[53]

On 18 March 2019, theUnited States Department of Energy andIntel announced the firstexaFLOP supercomputer would be operational atArgonne National Laboratory by the end of 2021. The computer, namedAurora, was delivered to Argonne by Intel andCray.^[54][55]

On 7 May 2019, The U.S. Department of Energy announced a contract withCray to build the "Frontier" supercomputer at Oak Ridge National Laboratory. Frontier, originally anticipated to be operational in 2021, was projected to be the world's most powerful computer, with a peak performance of greater than 1.5 exaflops.^[56]

Since June 2019, all TOP500 systems deliver a petaflop or more on theHigh Performance Linpack (HPL) benchmark, with the entry level to the list now at 1.022 petaflops.^[57]

In May 2022, the Frontier supercomputer broke theexascale barrier, completing more than aquintillion 64-bit floating point arithmetic calculations per second. Frontier clocked in at approximately 1.1 exaflops, beating out the previous record-holder,Fugaku.^[58][59] In June 2024, Aurora was the second computer on the TOP500 to post an exascale Rmax value, at 1.012 exaflops.^[60]

Since then, Frontier has been dethroned by El Capitan, hosted at Lawrence Livermore National Laboratory, with an HPL score of 1.742 exaflops.^[61]

Some major systems are not on the list. A prominent example is the NCSA'sBlue Waters which publicly announced the decision not to participate in the list^[62] because they do not feel it accurately indicates the ability of any system to do useful work.^[63]

Other organizations decide not to list systems for security and/or commercial competitiveness reasons. One such example is the National Supercomputing Center at Qingdao'sOceanLight supercomputer, completed in March 2021, which was submitted for, and won, theGordon Bell Prize. The computer is an exaflop computer, but was not submitted to the TOP500 list; the first exaflop machine submitted to the TOP500 list was Frontier. Analysts suspected that the reason the NSCQ did not submit what would otherwise have been the world's first exascale supercomputer was to avoid inflaming political sentiments and fears within the United States, in the context of the United States – China trade war.^[64] Similarly, government agencies like theNational Security Agency formerly submitted their devices to the TOP500, only to stop after 1998.^[65]

Additional purpose-built machines that are not capable or do not run the benchmark were not included, such asRIKEN MDGRAPE-3 andMDGRAPE-4.

A GoogleTensor Processing Unit v4 pod is capable of 1.1 exaflops of peak performance,^[66] while TPU v5p claims over 4 exaflops inBfloat16 floating-point format,^[67] however, these units are highly specialized to runmachine learning workloads and the TOP500 measures a specific benchmark algorithm using a specific numeric precision.

Tesla Dojo's primary unnamed cluster using 5,760NvidiaA100graphics processing units (GPUs) was touted byAndrej Karpathy in 2021 at the fourthInternational Joint Conference on Computer Vision and Pattern Recognition (CCVPR 2021) to be "roughly the number five supercomputer in the world"^[68] at approximately 81.6petaflops, based on scaling the performance of the NvidiaSelene supercomputer, which uses similar components.^[69]

In March 2024,Meta AI disclosed the operation of two datacenters with 24,576 H100 GPUs,^[70] which is almost 2x as on the Microsoft Azure Eagle (#3 as of September 2024), which could have made them occupy 3rd and 4th places in TOP500, but neither have been benchmarked. During company's Q3 2024earnings call in October,M. Zuckerberg disclosed usage of a cluster with over 100,000 H100s.^[71]

xAI Memphis Supercluster (also known as "Colossus") allegedly features 100,000 of the same H100 GPUs, which could have put it in the first place, but it is reportedly not in full operation due to power shortages.^[72]

After the onset ofUS-China Trade War, China has largely shrouded its newly online supercomputers and data centers in secrecy, opting out of reporting to the TOP500 list.^[4] This is partly driven by fears of being targeted by US sanctions placed on Chinese domestic suppliers.^[73][5]

IBM Roadrunner^[74] is no longer on the list (nor is any other using theCell coprocessor, orPowerXCell).

AlthoughItanium-based systems reached second rank in 2004,^[75][76] none now remain.

Similarly (non-SIMD-style)vector processors (NEC-based such as theEarth simulator that was fastest in 2002^[77]) have also fallen off the list. Also theSun Starfire computers that occupied many spots in the past now no longer appear.

The last non-Linux computers on the list – the twoAIX ones – running onPOWER7 (in July 2017 ranked 494th and 495th,^[78] originally 86th and 85th), dropped off the list in November 2017.

• The first edition of TOP500 to feature only 64-bit supercomputers was the 59th edition of TOP500, which was published in June 2022.

Wikimedia Commons has media related toTOP500.

• Computer science
• Computing
• Graph500
• Green500
• HPC Challenge Benchmark
• Instructions per second
• LINPACK benchmarks
• List of fastest computers
• List of the top supercomputers in the United States

• Official website
• LINPACK benchmarks at TOP500

• Supercomputers
• Supercomputer benchmarks
• Supercomputer sites
• Top lists

• CS1 Simplified Chinese-language sources (zh-hans)
• Articles with short description
• Short description is different from Wikidata
• Use dmy dates from November 2015
• All articles with unsourced statements
• Articles with unsourced statements from December 2023
• Articles with unsourced statements from March 2014
• Wikipedia articles in need of updating from August 2022
• All Wikipedia articles in need of updating
• Articles containing potentially dated statements from June 2022
• All articles containing potentially dated statements
• Articles with unsourced statements from October 2024
• Articles containing potentially dated statements from November 2017
• Articles containing potentially dated statements from November 2025
• Articles needing additional references from September 2019
• All articles needing additional references
• Articles containing potentially dated statements from May 2017
• Commons category link is on Wikidata
• Official website different in Wikidata and Wikipedia