Summit (supercomputer)

{{Infobox custom computer

| Image = Summit (supercomputer) logo 2017.svg

| Caption =

| Dates =

| Operators = IBM

| Sponsors = United States Department of Energy

| Location =

| Architecture = 9,216 POWER9 22-core CPUs
27,648 Nvidia Tesla V100 GPUs{{Cite web|url=https://www.ornl.gov/news/ornl-launches-summit-supercomputer|title=ORNL Launches Summit Supercomputer|date=June 8, 2018|website=www.ornl.gov}}

| Storage = 250 PB

| Speed = 200 petaFLOPS (peak)

| Power = 13 MW{{cite web |last1=Liu |first1=Zhiye |title=US Dethrones China With IBM Summit Supercomputer |url=https://www.tomshardware.com/news/us-supercomputer-china-top500-summit,37367.html |website=Tom's Hardware |accessdate=19 July 2018 |language=en |date=26 June 2018}}

| OS = Red Hat Enterprise Linux (RHEL){{cite web |last1=Kerner |first1=Sean Michael |title=IBM Unveils Summit, the World's Fastest Supercomputer (For Now) |url=https://www.serverwatch.com/server-news/ibm-unveils-summit-the-worlds-faster-supercomputer-for-now.html |website=Server Watch |accessdate=24 February 2020 |language=en |date=8 June 2018}}{{cite web |last1=Nestor |first1=Marius |title=Meet IBM Summit, World's Fastest and Smartest Supercomputer Powered by Linux |url=https://news.softpedia.com/news/meet-summit-world-s-fastest-and-smartest-supercomputer-powered-by-linux-521506.shtml |website=Softpedia News |accessdate=24 February 2020 |language=en |date=11 June 2018}}

| Space =

| Cost =

| ChartName = TOP500

| ChartPosition = 7 (1H2024)

| ChartDate =

| Purpose = Scientific research

| Legacy =

| Emulators =

| Website = {{URL|https://www.olcf.ornl.gov/olcf-resources/compute-systems/summit/}}

}}

File:Summit (supercomputer).jpg

File:POWER9TOP500Certificates.jpg wafer with TOP500 certificates for Summit and Sierra]]

Summit or OLCF-4 was a supercomputer developed by IBM for use at Oak Ridge Leadership Computing Facility (OLCF), a facility at the Oak Ridge National Laboratory, United States of America. It held the number 1 position on the TOP500 list from June 2018 to June 2020.{{cite web|url=https://www.nytimes.com/2018/06/08/technology/supercomputer-china-us.html|title=Move Over, China: U.S. Is Again Home to World's Speediest Supercomputer|last=Lohr|first=Steve|date=8 June 2018|newspaper=The New York Times|accessdate=19 July 2018}}{{cite web|url=https://www.top500.org/lists/top500/2022/11/|title=Top 500 List - November 2022|date=November 2022|website=TOP500|language=en|accessdate=13 April 2022}} As of June 2024, its LINPACK benchmark was clocked at 148.6 petaFLOPS.{{cite web|url=https://www.top500.org/lists/2022/11/|title=November 2022 {{!}} TOP500 Supercomputer Sites|website=TOP500|language=en|accessdate=13 April 2022}} Summit was decommissioned on November 15, 2024.{{cite web|url=https://docs.olcf.ornl.gov/systems/2024_olcf_system_changes.html |title=2024 Notable System Changes — OLCF User Documentation|language=en|date=31 October 2024 |accessdate=1 January 2025}}

As of November 2019, the supercomputer had ranked as the 5th most energy efficient in the world with a measured power efficiency of 14.668 gigaFLOPS/watt.{{cite web|url=https://www.top500.org/green500/lists/2019/11/|title=Green500 List - November 2019|website=TOP500|language=en|accessdate=7 April 2020}} Summit was the first supercomputer to reach exaflop (a quintillion operations per second) speed, on a non-standard metric, achieving 1.88 exaflops during a genomic analysis and is expected to reach 3.3 exaflops using mixed-precision calculations.{{cite web |last1=Holt |first1=Kris |title=The US again has the world's most powerful supercomputer |url=https://www.engadget.com/2018/06/08/summit-supercomputer-research-ai-oak-ridge/ |website=Engadget |date=8 June 2018 |accessdate=20 July 2018}}

History

The United States Department of Energy awarded a $325 million contract in November 2014 to IBM, Nvidia and Mellanox. The effort resulted in construction of Summit and Sierra. Summit is tasked with civilian scientific research and is located at the Oak Ridge National Laboratory in Tennessee. Sierra is designed for nuclear weapons simulations and is located at the Lawrence Livermore National Laboratory in California.{{cite web|url=http://www.cnet.com/news/ibm-nvidia-land-325-million-supercomputer-deal/|title=IBM, NVIDIA land $325M supercomputer deal|last=Shankland|first=Steven|date=14 September 2015|publisher=C{{!}}Net|accessdate=29 December 2015}}

Summit was estimated to cover {{convert|5600|sqft|m2}}{{Cite web | url=https://www.olcf.ornl.gov/wp-content/uploads/2018/06/Summit_bythenumbers_FIN-1.pdf | title=America’s most powerful supercomputer is a machine for scientific discovery | website=www.olcf.ornl.gov}} and require {{convert|219|km|mi}} of cabling,{{cite news|last1=Alcorn|first1=Paul|title=Regaining America's Supercomputing Supremacy With The Summit Supercomputer|url=http://www.tomshardware.com/news/summit-supercomputer-nvidia-ibm-power9-volta,35962.html|accessdate=20 November 2017|publisher=Tom's Hardware|date=20 November 2017}} and was designed to be used for research in diverse fields such as cosmology, medicine, and climatology.{{cite web|url=http://www.pcworld.com/article/2947792/ibm-nvidia-rev-hpc-engines-in-nextgen-supercomputer-push.html|title=IBM, NVIDIA rev HPC engines in next-gen supercomputer push|last=Noyes|first=Katherine|date=16 March 2015|publisher=PC World|accessdate=29 December 2015}}

In 2015, the project called Collaboration of Oak Ridge, Argonne and Lawrence Livermore (CORAL) included a third supercomputer named Aurora and was planned for installation at Argonne National Laboratory.{{cite web|url=http://www.eetimes.com/document.asp?doc_id=1326359|title=IBM vs. Intel in Supercomputer Bout|last=R. Johnson|first=Colin|date=15 April 2015|publisher=EE Times|accessdate=29 December 2015}} By 2018, Aurora was re-engineered with completion anticipated in 2021 as an exascale computing project along with Frontier and El Capitan to be completed shortly thereafter.{{cite web |last1=Morgan |first1=Timothy Prickett |title=Bidders Off And Running After $1.8 Billion DOE Exascale Super Deals |url=https://www.nextplatform.com/2018/04/09/bidders-off-and-running-after-1-8-billion-doe-exascale-supercomputer-deals/ |website=The Next Platform |accessdate=20 July 2018 |date=9 April 2018}} Aurora was completed in late 2022.{{Cite web|last=Hemsoth|first=Nicole|date=2021-09-23|title=A Status Check on Global Exascale Ambitions|url=https://www.nextplatform.com/2021/09/23/a-status-check-on-global-exascale-ambitions/|access-date=2021-10-15|website=The Next Platform|language=en-US}}

Uses

The Summit supercomputer was built for research in energy, artificial intelligence, human health, and other areas.{{cite web|url=https://www.olcf.ornl.gov/summit/|title=Introducing Summit|accessdate=24 December 2019}} It has been used in earthquake simulation, extreme weather simulation, materials science, genomics, and predicting the lifetime of neutrinos.{{cite web|url=https://www.hpcwire.com/2018/09/20/summit-supercomputer-is-already-making-its-mark-on-science/|title=Summit Supercomputer is Already Making its Mark on Science|date=20 September 2018|accessdate=5 August 2020}}

Design

Each of its 4,608 nodes consist of 2 IBM POWER9 CPUs, 6 Nvidia Tesla GPUs,{{cite web |title=The most powerful computers on the planet - Summit and Sierra |url=https://www.ibm.com/thought-leadership/summit-supercomputer/ |website=IBM |date=6 June 2018 |accessdate=4 April 2019}} with over 600 GB of coherent memory (96 GB HBM2 plus 512 GB DDR4) which is addressable by all CPUs and GPUs, plus 800 GB of non-volatile RAM that can be used as a burst buffer or as extended memory.{{Cite web|url=http://hothardware.com/news/nvidia-12nm-finfet-volta-gpu-architecture-replacing-pascal-2017|title=NVIDIA 12nm FinFET Volta GPU Architecture Reportedly Replacing Pascal In 2017|last=Lilly|first=Paul|date=January 25, 2017|website=HotHardware}} The POWER9 CPUs and Nvidia Volta GPUs are connected using Nvidia's high speed NVLink. This allows for a heterogeneous computing model.{{Cite web|url=http://www.teratec.eu/actu/calcul/Nvidia_Coral_White_Paper_Final_3_1.pdf|title=Summit and Sierra Supercomputers: An Inside Look at the U.S. Department of Energy's New Pre-Exascale Systems|date=November 1, 2014}}

To provide a high rate of data throughput, the nodes are connected in a non-blocking fat-tree topology using a dual-rail Mellanox EDR InfiniBand interconnect for both storage and inter-process communications traffic, which delivers both 200 Gbit/s bandwidth between nodes and in-network computing acceleration for communications frameworks such as MPI and SHMEM/PGAS.

The storage for Summit {{Cite report |url=https://www.osti.gov/biblio/1619016 |title=End-to-end I/O portfolio for the summit supercomputing ecosystem |last1=Oral |first1=Sarp |last2=Vazhkudai |first2=Sudharshan |date=2019-11-01 |publisher=Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States) |language=English |last3=Wang |first3=Feiyi |last4=Zimmer |first4=Christopher |last5=Brumgard |first5=Christopher |last6=Hanley |first6=Jesse |last7=Markomanolis |first7=George |last8=Miller |first8=Ross |last9=Leverman |first9=Dustin B.|osti=1619016 }} has a fast in-system layer and a center-wide parallel filesystem layer. The in-system layer is optimized for fast storage with SSDs on each node, while the center-wide parallel file system provides easy to access data stored on hard drives. The two layers work together seamlessly so users do not have to differentiate their storage needs. The center-wide parallel file system is GPFS (IBM Storage Scale). It provides 250PB of storage. The cluster delivers 2.5 TB/s of single stream read peak throughput and 1 TB/s of 1M file throughput. It was one of the first supercomputers that also required extremely fast metadata performance to support AI/ML workloads exemplified by the 2.6M 32k file creates per second it delivers.