топ 500 суперкомпьютеров 2020 список
June 2020
The 55th edition of the TOP500 saw some significant additions to the list, spearheaded by a new number one system from Japan. The latest rankings also reflect a steady growth in aggregate performance and power efficiency.
The new top system, Fugaku, turned in a High Performance Linpack (HPL) result of 415.5 petaflops, besting the now second-place Summit system by a factor of 2.8x. Fugaku, is powered by Fujitsu’s 48-core A64FX SoC, becoming the first number one system on the list to be powered by ARM processors. In single or further reduced precision, which are often used in machine learning and AI applications, Fugaku’s peak performance is over 1,000 petaflops (1 exaflops). The new system is installed at RIKEN Center for Computational Science (R-CCS) in Kobe, Japan.
Number two on the list is Summit, an IBM-built supercomputer that delivers 148.8 petaflops on HPL. The system has 4,356 nodes, each equipped with two 22-core Power9 CPUs, and six NVIDIA Tesla V100 GPUs. The nodes are connected with a Mellanox dual-rail EDR InfiniBand network. Summit is running at Oak Ridge National Laboratory (ORNL) in Tennessee and remains the fastest supercomputer in the US.
At number three is Sierra, a system at the Lawrence Livermore National Laboratory (LLNL) in California achieving 94.6 petaflops on HPL. Its architecture is very similar to Summit, equipped with two Power9 CPUs and four NVIDIA Tesla V100 GPUs in each of its 4,320 nodes. Sierra employs the same Mellanox EDR InfiniBand as the system interconnect.
Sunway TaihuLight, a system developed by China’s National Research Center of Parallel Computer Engineering & Technology (NRCPC) drops to number four on the list. The system is powered entirely by Sunway 260-core SW26010 processors. Its HPL mark of 93 petaflops has remained unchanged since it was installed at the National Supercomputing Center in Wuxi, China in June 2016.
At number five is Tianhe-2A (Milky Way-2A), a system developed by China’s National University of Defense Technology (NUDT). Its HPL performance of 61.4 petaflops is the result of a hybrid architecture employing Intel Xeon CPUs and custom-built Matrix-2000 coprocessors. It is deployed at the National Supercomputer Center in Guangzhou, China.
A new system on the list, HPC5, captured the number six spot, turning in an HPL performance of 35.5 petaflops. HPC5 is a PowerEdge system built by Dell and installed by the Italian energy firm Eni S.p.A, making it the fastest supercomputer in Europe. It is powered by Intel Xeon Gold processors and NVIDIA Tesla V100 GPUs and uses Mellanox HDR InfiniBand as the system network.
Another new system, Selene, is in the number seven spot with an HPL mark of 27.58 petaflops. It is a DGX SuperPOD, powered by NVIDIA’s new “Ampere” A100 GPUs and AMD’s EPYC “Rome” CPUs. Selene is installed at NVIDIA in the US. It too uses Mellanox HDR InfiniBand as the system network.
Frontera, a Dell C6420 system installed at the Texas Advanced Computing Center (TACC) in the US is ranked eighth on the list. Its 23.5 HPL petaflops is achieved with 448,448 Intel Xeon cores.
The second Italian system in the top 10 is Marconi-100, which is installed at the CINECA research center. It is powered by IBM Power9 processors and NVIDIA V100 GPUs, employing dual-rail Mellanox EDR InfiniBand as the system network. Marconi-100’s 21.6 petaflops earned it the number nine spot on the list.
Rounding out the top 10 is Piz Daint at 21.2 petaflops, a Cray XC50 system installed at the Swiss National Supercomputing Centre (CSCS) in Lugano, Switzerland. It is equipped with Intel Xeon processors and NVIDIA P100 GPUs.
TOP 10 Sites for June 2020
For more information about the sites and systems in the list, click on the links or view the complete list.
November 2020
The 56th edition of the TOP500 saw the Japanese Fugaku supercomputer solidify its number one status in a list that reflects a flattening performance growth curve. Although two new systems managed to make it into the top 10, the full list recorded the smallest number of new entries since the project began in 1993.
The entry level to the list moved up to 1.32 petaflops on the High Performance Linpack (HPL) benchmark, a small increase from 1.23 petaflops recorded in the June 2020 rankings. In a similar vein, the aggregate performance of all 500 systems grew from 2.22 exaflops in June to just 2.43 exaflops on the latest list. Likewise, average concurrency per system barely increased at all, growing from 145,363 cores six months ago to 145,465 cores in the current list.
There were, however, a few notable developments in the top 10, including two new systems, as well as a new highwater mark set by the top-ranked Fugaku supercomputer. Thanks to additional hardware, 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, 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 represents the first benchmark measurements above one exaflop for any precision on any type of hardware.
Here is a brief rundown of current top 10 systems:
A new supercomputer, known as the JUWELS Booster Module, debuts at number seven on the list. The Atos-built BullSequana machine was recently installed at the Forschungszentrum Jülich (FZJ) in Germany. It is part of a modular system architecture and a second Xeon based JUWELS Module is listed separately on the TOP500 at position 44. These modules are integrated by using the ParTec Modulo Cluster Software Suite. The Booster Module uses AMD EPYC processors with NVIDIA A100 GPUs for acceleration similar to the number five Selene system. Running by itself the JUWELS Booster Module was able to achieve 44.1 HPL petaflops, which makes it the most powerful system in Europe
TOP 10 Sites for November 2020
For more information about the sites and systems in the list, click on the links or view the complete list.
Опубликован 55-й выпуск рейтинга топ-500 самых высокопроизводительных систем мира, там появился новый лидер Fugaku
22 июня 2020 года опубликован 55-й выпуск рейтинга топ-500 самых высокопроизводительных систем мира. В списке суперкомпьютеров мира произошла смена лидера. Вместо американского Summit (IBM) на первом месте теперь японский кластер Fugaku на процессорах ARM A64FX 48C 2.2 GHz. Заявленная пиковая производительность Fugaku составляет 513.8 петафлопс, что в 2,5 раза больше, чем у суперкомпьютера Summit на IBM Power9.
Суперкомпьютер Fugaku был спроектирован и собран в японском институте физико-химических исследований RIKEN. Его разработка была инициирована Министерством образования, культуры, спорта, науки и технологий Японии в 2014 году. Инженерам и специалистам по высокопроизводительным системам была поставлена цель — создать флагманский суперкомпьютер следующего поколения (эксафлопсный суперкомпьютер). Его официальный запуск в промышленную эксплуатацию будет в 2021 году, к этому времени Fugaku будет еще модернизирован.
Основные вехи проектирования, разработки и сборки суперкомпьютера Fugaku.
В настоящее время специалисты института RIKEN занимаются усовершенствованием и доработкой систем и элементов Fugaku.
Суперкомпьютера Fugaku состоит из 158 976 узлов на базе SoC Fujitsu A64FX (в каждом из них на борту 48-ядерный CPU Armv8.2-A SVE (512 bit SIMD) с тактовой частотой 2.2 ГГц). В сумме кластер Fugaku насчитывает более 7 млн процессорных ядер, около 5 ПБ ОЗУ и 150 ПБ общего хранилища на базе ФС Lustre. В качестве операционной системы в Fugaku используется Red Hat Enterprise Linux.
SoC Fujitsu A64FX уникален тем, что он является первым процессором, в котором объединена поддержка памяти HBM2 и векторных расширений архитектуры Arm (Scalable Vector Extensions или SVE). Использование HBM2 обеспечивает A64FX теоретическую пропускную способность памяти более 1 ТБ/с, а поддержка Arm SVE повышает производительность в задачах искусственного интеллекта и аналитики. Японские инженеры уже несколько лет прорабатывают решения для применения этих особенности A64FX как для исследований в области ядерной физики так и в других отраслях науки и промышленности, использую преимущества систем на базе Arm — высокая степень распараллеливания, низкое энергопотребление и высокая надежность.
Кстати, у проекта Fujitsu A64FX есть свой открытый репозиторий в GitHub, где размещена документация и примеры для разработчиков.
Рейтинг топ-500 суперкомпьютеров мира составляется с 1993 года два раза в год (в июне и ноябре) специалистами и учеными из США и Германии. Минимальный порог для попадания в него в этом году достиг 1,702 петафлопс. В новой версии рейтинга топ-500 суперкомпьютеров мира Китаю принадлежат 226 систем, 114 относятся к США и 29 расположены в Японии.
Количество российских суперкомпьютеров в рейтинге топ-500 самых высокопроизводительных систем мира уменьшилось. Теперь там только две отечественные высокопроизводительные системы — на 36 месте SberCloud (пиковая производительность 8,789 петафлопс) и на 131 месте Lomonosov 2 (пиковая производительность 4,947 петафлопс). В этом году суперкомпьютер ФГБУ ГВЦ Росгидромета (пиковая производительность 1,3 петафлопс) выбыл из нового списка топ-500 самых высокопроизводительных систем мира.
Ранее в конце марта 2020 года был обновлен рейтинг 50 самых производительных суперкомпьютеров СНГ (редакция №32 от 31.03.2020), в котором достаточно долгое время не появляются новые высокопроизводительные системы, исключение — суперкомпьютер Christofari, разработанный облачной платформой «Сбербанка» SberCloud. Все 50 суперкомпьютеров СНГ в качестве основных процессоров используют чипы Intel и AMD, также среди них 28 гибридных систем, использующих для вычислений ускорители.
На данный момент суперкомпьютеров на базе отечественной архитектуры «Эльбрус» в списке топ-50 суперкомпьютеров СНГ нет, хотя ранее в концерне «Автоматика» госкорпорации Ростех анонсировали появление первого суперкомпьютера на 8-ядерных микропроцессорах «Эльбрус-8С», который предназначен для выполнения высокопроизводительных вычислений, обработки больших объемов данных и решения задач с повышенными требованиями к информационной безопасности.
22 июня 2020 года на конференции WWDC 2020 компания Apple официально объявила о переходе на собственные процессоры на базе ARM в компьютерах Mac в будущем. Специалисты предполагают, что переход Apple на ARM вызван замедлением роста производительности Intel. Инженеры Apple тестировали чипы на базе ARM в компьютерах Mac и обнаружили значительное повышение производительности по сравнению с альтернативами Intel.
TOP500 News
GREEN500: Trend of steady progress with no big step toward newer technologies.
June 28, 2021
Although there was a trend of steady progress in the Green500, nothing has indicated a big step toward newer technologies.
The system to snag the No. 1 spot for the Green500 was MN-3 from Preferred Networks in Japan. Knocked from the top of the last list by NVIDIA DGX SuperPOD in the US, MN-3 is back to reclaim its crown. This system relies on the MN-Core chip, an accelerator optimized for matrix arithmetic, as well as a Xeon Platinum 8260M processor. MN-3 achieved a 29.70 gigaflops/watt power-efficiency and has a TOP500 ranking of 337.
Report on the Fujitsu Fugaku system by Jack Dongarra
June 22, 2020
Detailed report on the Fujitsu Fugaku system.
News Feed
Amazon Is The Flywheel, AWS Is The Cash Register
The coronavirus pandemic giveth to Amazon retail business and its Amazon Web Services cloud business, and the pandemic taketh away from the Amazon retail business. …
LLNL Researchers Garner 3 Awards Among Top 100 Industrial Inventions
October 29, 2021 — Lawrence Livermore National Laboratory (LLNL) scientists and engineers have collected three awards among the top 100 industrial inventions worldwide. The trade journal R&D World Magazine announced the winners of the awards, often called the “Oscars of invention,” during a three-day virtual ceremony – Oct. 19-21 – and on the magazine’s website. With this […]
HPE’s HPC Leader Justin Hotard on the Broader Impacts and Implications of Exascale
In this interview with Hewlett Packard Enterprise’s SVP/GM of High Performance Computing & Mission Critical Solutions Justin Hotard, he discusses the new world to be enabled by exascale supercomputing. Among the topics he delves into: why exascale is applicable to more market sectors (read: the enterprise) than some may think and how the Cray programming environment is facilitating supercomputing’s rapidly evolving heterogeneous super power: HPC combined with AI.
Innovative Chip Built by U of Copenhagen Physicists Resolves Quantum Headache
Oct. 29, 2021 — Quantum physicists at the University of Copenhagen are reporting an international achievement for Denmark in the field of quantum technology. By simultaneously operating multiple spin qubits on the same quantum chip, they surmounted a key obstacle on the road to the supercomputer of the future. The result bodes well for the […]
Vlodomyr Kindratenko Named Director of the Center for Artificial Intelligence Innovation at NCSA
Oct. 29, 2021 — NCSA is pleased to announce the appointment of Dr. Vlodomyr Kindratenko as Director of the Center for Artificial Intelligence Innovation (CAII). In this new role, he will be responsible for providing the overall leadership, oversight and management of the center, including developing partnerships and projects at regional and national levels, and overseeing […]
The Accelerated Path To Petabyte-Scale Graph Databases
Database acceleration using specialized co-processors is nothing new. Just to give a few examples, data warehouses running on the Netezza platform, owned by IBM for more than a decade now, uses a custom and parallelized PostgreSQL database matched to FPGA acceleration for database and storage routines. …
Sponsored Article
PRACE Software Strategy for European Exascale Systems
Sept. 1, 2021
Building on the successful implementation of the Partnership for Advanced Computing in Europe (PRACE), the European Commission (EC) has increased its efforts to develop a world-class supercomputing ecosystem in Europe. The EC, EuroHPC Joint Undertaking (JU) and EU Member States have made significant investments in European petascale and pre-exascale infrastructure, have put exascale supercomputers on the roadmap, and are actively exploring new post-exascale architectures. The return on investment will be directly linked to the productivity of end-users in academia, in industry, and in the public sector. Key to this productivity is an ecosystem of user-oriented software: scientific applications and workflows …
The List
06/2021 Highlights
The only new entry in the Top10 is the Perlmutter system at NERSC at the DOE Lawrence Berkeley National Laboratory. It is based on the HPE Cray “Shasta” platform and a heterogeneous system with both GPU-accelerated and CPU-only nodes. Perlmutter achieved 64.6 Pflop/s which put it at No. 5 in the new list.
TOP500
June 2021
The 57th TOP500 List was published June 28, 2021 in Virtual.
November 2020
The 56th TOP500 List was published Nov. 16, 2020 in Virtual.
June 2020
The 55th TOP500 List was published June 22, 2020 in Frankfurt, Germany.
November 2019
The 54th TOP500 List was published Nov. 18, 2019 in Denver, CO.
June 2019
The 53rd TOP500 List was published June 18, 2019 in Frankfurt, Germany.
November 2018
The 52nd TOP500 List was published Nov. 13, 2018 in Dallas, TX.
June 2018
The 51st TOP500 List was published June 25, 2018 in Frankfurt, Germany.
November 2017
The 50th TOP500 List was published Nov. 14, 2017 in Denver, CO.
June 2017
The 49th TOP500 List was published June 20, 2017 in Frankfurt, Germany.
November 2016
The 48th TOP500 List was published Nov. 13, 2016 in Salt Lake City, Utah.
June 2016
The 47th TOP500 List was published June 20, 2016 in Frankfurt, Germany.
November 2015
The 46th TOP500 List was published Nov. 17, 2015 in Austin, TX.
June 2015
The 45th TOP500 List was published June 14, 2015 in Frankfurt, Germany.
November 2014
The 44th TOP500 List was published Nov. 18, 2014 in New Orleans, LA.
June 2014
The 43rd TOP500 List was published June 23, 2014 in Leipzig, Germany.
November 2013
The 42nd TOP500 List was published Nov. 17, 2013 in Denver, CO.
June 2013
The 41st TOP500 List was published June 17, 2013 in Leipzig, Germany.
November 2012
The 40th TOP500 List was published Nov. 13, 2012 in Salt Lake City, Utah.
June 2012
The 39th TOP500 List was published June 18, 2012 in Hamburg, Germany.
November 2011
The 38th TOP500 List was published Nov. 13, 2011 in Seattle, WA.
June 2011
The 37th TOP500 List was published June 20, 2011 in Hamburg, Germany.
November 2010
The 36th TOP500 List was published Nov. 16, 2010 in New Orleans, LA.
June 2010
The 35th TOP500 List was published June 1, 2010 in Hamburg, Germany.
November 2009
The 34th TOP500 List was published Nov. 16, 2009 in Portland, Oregon.
June 2009
The 33rd TOP500 List was published June 23, 2009 in Hamburg, Germany.
November 2008
June 2008
The 31st TOP500 List was published June 18, 2008 in Dresden, Germany.
November 2007
The 30th TOP500 List was published Nov. 20, 2007 in Reno, Nevada.
June 2007
The 29th TOP500 List was published June 27, 2007 in Dresden, Germany.
November 2006
The 28th TOP500 List was published Nov. 15, 2006 in Tampa, FL.
June 2006
The 27th TOP500 List was published June 27, 2006 in Dresden, Germany.
November 2005
The 26th TOP500 List was published Nov. 16, 2005 in Seattle, WA.
June 2005
The 25th TOP500 List was published June 22, 2005 in Heidelberg, Germany.
November 2004
The 24th TOP500 List was published Nov. 8, 2004 in Pittsburgh, PA.
June 2004
The 23rd TOP500 List was published June 22, 2004 in Heidelberg, Germany.
November 2003
The 22nd TOP500 List was published Nov. 15, 2003 in Phoenix, AZ.
June 2003
The 21st TOP500 List was published June 30, 2003 in Heidelberg, Germany.
November 2002
The 20th TOP500 List was published Nov. 16, 2002 in Baltimore, MD.
June 2002
The 19th TOP500 List was published June 30, 2002 in Heidelberg, Germany.
November 2001
The 18th TOP500 List was published Nov. 10, 2001 in Denver, CO.
June 2001
The 17th TOP500 List was published June 30, 2001 in Heidelberg, Germany.
November 2000
The 16th TOP500 List was published Nov. 4, 2000 in Dallas, TX.
June 2000
The 15th TOP500 List was published June 30, 2000 in Mannheim, Germany.
November 1999
The 14th TOP500 List was published Nov. 13, 1999 in Portland, OR.
June 1999
The 13th TOP500 List was published June 30, 1999 in Mannheim, Germany.
November 1998
The 12th TOP500 List was published Nov. 7, 1998 in Orlando, FL.
June 1998
The 11th TOP500 List was published June 30, 1998 in Mannheim, Germany.
November 1997
The 10th TOP500 List was published Nov. 15, 1997 in San Jose, CA.
June 1997
The 9th TOP500 List was published June 30, 1997 in Mannheim, Germany.
November 1996
The 8th TOP500 List was published Nov. 17, 1996 in Pittsburgh, PA.
June 1996
The 7th TOP500 List was published June 30, 1996 in Mannheim, Germany.
November 1995
The 6th TOP500 List was published Dec. 4, 1995 in San Diego, CA.
June 1995
The 5th TOP500 List was published June 30, 1995 in Mannheim, Germany.
November 1994
The 4th TOP500 List was published Nov. 13, 1994 in Washington, D.C..
June 1994
The 3rd TOP500 List was published June 30, 1994 in Mannheim, Germany.
November 1993
The 2nd TOP500 List was published Nov. 15, 1993 in Portland, OR.
June 1993
The 1st TOP500 List was published June 30, 1993 in Mannheim, Germany.