Exascale computing refers tocomputing systems capable of calculating at least 1018IEEE 754 double precision (64-bit) operations (multiplications and/or additions) per second (exaFLOPS)";[1] it is a measure ofsupercomputer performance.
Exascale computing is a significant achievement incomputer engineering: primarily, it allows improved scientific applications and better prediction accuracy in domains such asweather forecasting,climate modeling andpersonalised medicine.[2] Exascale also reaches the estimated processing power of thehuman brain at the neural level, a target of the now defunctHuman Brain Project.[3] There has been a race to be the first country to build an exascale computer, typically ranked in theTOP500 list.[4][5][6][7]
In 2022, the world's first public exascale computer,Frontier, was announced.[8] As of November 2024[update], Lawrence Livermore National Laboratory'sEl Capitan is the world's fastest exascale supercomputer.[9]
A new exascale supercomputer,JUPITER,[10] was inaugurated inGermany in 2025. Although it is the 4th in the world ranking at the moment, it has the number‑one position on theGreen500 ranking, because the system runs entirely on renewable energy and features cutting-edge cooling and energy reuse, making it the world's most energy-efficient supercomputer.[11]
Floating point operations per second (FLOPS) are one measure ofcomputer performance. FLOPS can be recorded in different measures of precision, however the standard measure (used by theTOP500 supercomputer list) uses 64 bit (double-precision floating-point format) operations per second using theHigh Performance LINPACK (HPLinpack)benchmark.[12][1]
Whilst a distributed computing system had broken the 1 exaFLOPS barrier beforeFrontier, the metric typically refers to single computing systems. Supercomputers had also previously broken the 1 exaFLOPS barrier using alternative precision measures; again these do not meet the criteria for exascale computing using the standard metric.[1] It has been recognised that HPLinpack may not be a good general measure of supercomputer utility in real world application, however it is the common standard for performance measurement.[13][14]
It has been recognized that enabling applications to fully exploit capabilities of exascale computing systems is not straightforward.[15] Developing data-intensive applications over exascale platforms requires the availability of new and effective programming paradigms and runtime systems.[16] TheFolding@home project, the first to break this barrier, relied on a network of servers sending pieces of work to hundreds of thousands of clients using aclient–server modelnetwork architecture.[17][18]
The firstpetascale (1015 FLOPS) computer entered operation in 2008.[19] At asupercomputing conference in 2009,Computerworld projected exascale implementation by 2018.[20] In June 2014, the stagnation of theTop500 supercomputer list had observers question the possibility of exascale systems by 2020.[21]
Although exascale computing was not achieved by 2018, in the same year theSummit OLCF-4 supercomputer performed 1.8×1018 calculations per second using an alternative metric whilst analysing genomic information.[22] The team performing this won theGordon Bell Prize at the 2018ACM/IEEE Supercomputing Conference.[citation needed]
The exaFLOPS barrier was first broken in March 2020 by thedistributed computing networkFolding@homecoronavirus research project.[23][18][24][25][26]
In June 2020[27] the Japanese supercomputerFugaku achieved 1.42 exaFLOPS using the alternative HPL-AI benchmark.
In 2022, the world's first public exascale computer,Frontier, was announced, achieving an Rmax of 1.102 exaFLOPS in June 2022.[8] As of November 2024[update], the world's fastest supercomputer isEl Capitan at 1.742 exaFLOPS.[9]
In 2008, twoUnited States of America governmental organisations within theUS Department of Energy, theOffice of Science and theNational Nuclear Security Administration, provided funding to the Institute for Advanced Architectures for the development of an exascale supercomputer;Sandia National Laboratory and theOak Ridge National Laboratory were also to collaborate on exascale designs.[28] The technology was expected to be applied in various computation-intensive research areas, includingbasic research,engineering,earth science,biology,materials science, energy issues, and national security.[29]
In January 2012,Intel purchased theInfiniBand product line fromQLogic for US$125 million in order to fulfill its promise of developing exascale technology by 2018.[30]
By 2012, the United States had allotted $126 million for exascale computing development.[31]
In February 2013,[32] theIntelligence Advanced Research Projects Activity started the Cryogenic Computer Complexity (C3) program, which envisions a new generation ofsuperconductingsupercomputers that operate at exascale speeds based onsuperconducting logic. In December 2014 it announced a multi-year contract with IBM, Raytheon BBN Technologies and Northrop Grumman to develop the technologies for the C3 program.[33]
On 29 July 2015,Barack Obama signed an executive order creating aNational Strategic Computing Initiative calling for the accelerated development of an exascale system and funding research into post-semiconductor computing.[34] The Exascale Computing Project (ECP) hopes to build an exascale computer by 2021.[35]
On 18 March 2019, theUnited States Department of Energy andIntel announced the first exaFLOPS supercomputer would be operational atArgonne National Laboratory by late 2022. The computer, namedAurora is to be delivered to Argonne by Intel andCray (now Hewlett Packard Enterprise), and is expected to use Intel Xe GPGPUs alongside a future Xeon Scalable CPU, and cost US$600 Million.[36][37]
On 7 May 2019, the U.S. Department of Energy announced a contract with Cray (now Hewlett Packard Enterprise) to build theFrontier supercomputer at Oak Ridge National Laboratory. Frontier is anticipated to be fully operational in 2022[38] and, with a performance of greater than 1.5 exaFLOPS, should then be the world's most powerful computer.[39]
On 4 March 2020, the U.S. Department of Energy announced a contract withHewlett Packard Enterprise and AMD to build theEl Capitan supercomputer at a cost of US$600 million, to be installed at theLawrence Livermore National Laboratory (LLNL). It is expected to be used primarily (but not exclusively) for nuclear weapons modeling. El Capitan was first announced in August 2019, when the DOE and LLNL revealed the purchase of a Shasta supercomputer from Cray. El Capitan will be operational in early 2023 and have a performance of 2 exaFLOPS. It will use AMD CPUs and GPUs, with 4 Radeon Instinct GPUs per EPYC Zen 4 CPU, to speed up artificial intelligence tasks. El Capitan should consume around 40 MW of electric power.[40][41]
In May 2022, the United States had its first exascale supercomputer, Frontier. In June 2024, Argonne National Laboratory'sAurora became the country's second exascale computer, followed five months later by El Capitan becoming operational.
In Japan, in 2013, theRIKEN Advanced Institute for Computational Science began planning an exascale system for 2020, intended to consume less than 30 megawatts.[42] In 2014,Fujitsu was awarded a contract by RIKEN to develop a next-generation supercomputer to succeed theK computer. The successor is calledFugaku, and aims to have a performance of at least 1 exaFLOPS, and be fully operational in 2021. In 2015, Fujitsu announced at theInternational Supercomputing Conference that this supercomputer would use processors implementing theARMv8 architecture with extensions it was co-designing withARM Limited.[43] It was partially put into operation in June 2020[27] and achieved 1.42 exaFLOPS (fp16 with fp64 precision) in HPL-AI benchmark making it the first ever supercomputer that achieved 1 exaFLOPS.[44] Named after Mount Fuji, Japan's tallest peak, Fugaku retained the No. 1 ranking on the Top 500 supercomputer calculation speed ranking announced on November 17, 2020, reaching a calculation speed of 442 quadrillion calculations per second, or 0.442 exaFLOPS.[45]
As of June 2022, China had two of the Top Tenfastest supercomputers in the world. According to the national plan for the next generation of high performance computers and the head of the school of computing at the National University of Defense Technology (NUDT), China was supposed to develop an exascale computer during the 13th Five-Year-Plan period (2016–2020) which would enter service in the latter half of 2020.[46] The government of Tianjin Binhai New Area, NUDT and the National Supercomputing Center in Tianjin are working on the project. AfterTianhe-1 andTianhe-2, the exascale successor is planned to be named Tianhe-3. As of 2023, China is reported to have two operational exascale computers; Tianhe-3 (Xingyi)[47] and Sunway OceanLight, with a third being built. Neither are on the Top500.[48][49]
In 2011, several projects aiming at developing technologies and software for exascale computing were started in the European Union. The CRESTA project (Collaborative Research into Exascale Systemware, Tools and Applications),[50] the DEEP project (Dynamical ExaScale Entry Platform),[51] and the project Mont-Blanc.[52] A major European project based on exascale transition is the MaX (Materials at the Exascale) project.[53] The Energy oriented Centre of Excellence (EoCoE) exploits exascale technologies to support carbon-free energy research and applications.[54]
In 2015, the Scalable, Energy-Efficient, Resilient and Transparent Software Adaptation (SERT) project, a major research project between theUniversity of Manchester and the STFCDaresbury Laboratory inCheshire, was awarded c. £1million from the United Kingdom's Engineering and Physical Sciences Research Council. The SERT project was due to start in March 2015. It will be funded by EPSRC under the Software for the Future II programme, and the project will partner with the Numerical Analysis Group (NAG), Cluster Vision and the Science and Technology Facilities Council (STFC).[55]
On 28 September 2018, theEuropean High-Performance Computing Joint Undertaking (EuroHPC JU) was formally established by the EU. The EuroHPC JU aims to build an exascale supercomputer by 2022/2023. The EuroHPC JU will be jointly funded by its public members with a budget of around €1 billion. The EU's financial contribution is €486 million.[56][57] In 2025, theJUPITER supercomputer hosted byForschungszentrum Jülich is completed,[58] marking the first exascale supercomputer on the TOP500 list outside the United States. A second exascale supercomputer, Alice Recoque, is announced in 2023; the supercomputer will be hosted byGENCI and located at Très Grand Centre de calcul, in theCEA site atBruyères-le-Châtel, France.[59]
In March 2023 the government of the United Kingdom announced it would invest £900 million in the development of an exascale computer.[60] This project was axed in August 2024.[61]
In June 2017,Taiwan'sNational Center for High-Performance Computing initiated the effort towards designing and building the first Taiwanese exascale supercomputer by funding construction of a new intermediary supercomputer based on a full technology transfer fromFujitsu corporation ofJapan, which is currently building the fastest and most powerfulA.I. based supercomputer inJapan.[62][63][64][65][66]Additionally, numerous other independent efforts have been made in Taiwan with the focus on the rapid development of exascale supercomputing technology, such asFoxconn Corporation which recently designed and built the largest and fastest supercomputer in all of Taiwan. This newFoxconn supercomputer is designed to serve as a stepping stone in research and development towards the design and building of a state of the art exascale supercomputer.[67][68][69][70]
In 2012, the Indian Government proposed to commit US$2.5 billion to supercomputing research during the12th five-year plan period (2012–2017). The project was to be handled byIndian Institute of Science (IISc),Bangalore.[71] Additionally, it was later revealed that India plans to develop a supercomputer with processing power in theexaFLOPS range.[72] It will be developed byC-DAC within the subsequent five years of approval.[73] These supercomputers will use indigenously developed microprocessors by C-DAC in India.[74] In 2023, in a presentation by CDAC, it plans to have a indigenously developed exascale supercomputer named Param Shankh. The Param Shankh will be powered by an indigenous 96-core, ARM architecture-based processor which has been nicknamed AUM (ॐ).[75]