Skip to main content

Summit is the next leap in leadership-class computing systems for open science. With Summit we will be able to address, with greater complexity and higher fidelity, questions concerning who we are, our place on earth, and in our universe.

Summit, launched in 2018, delivers 8 times the computational performance of Titan’s 18,688 nodes, using only 4,608 nodes. Like Titan, Summit has a hybrid architecture, and each node contains multiple IBM POWER9 CPUs and NVIDIA Volta GPUs all connected together with NVIDIA’s high-speed NVLink. Each node has over half a terabyte of coherent memory (high bandwidth memory + DDR4) addressable by all CPUs and GPUs plus 800GB of non-volatile RAM that can be used as a burst buffer or as extended memory. To provide a high rate of I/O throughput, the nodes are connected in a non-blocking fat-tree using a dual-rail Mellanox EDR InfiniBand interconnect.

Summit allows researchers in all fields of science unprecedented access to solving some of the world’s most pressing challenges.

System Support

The System User Guide is the definitive source of information about Summit, and details everything from connecting to running complex workflows. Please direct questions about Summit and its usage to the OLCF User Assistance Center by emailing [email protected].

System Specifications

Processor:IBM POWER9™ (2/node)
GPUs:27,648 NVIDIA Volta V100s (6/node)
Nodes: 4,608
Node Performance:42TF
Memory/node:512GB DDR4 + 96GB HBM2
NV Memory/node:1600GB
Total System Memory:>10PB DDR4 + HBM + Non-volatile
Interconnect Topology:Mellanox EDR 100G InfiniBand, Non-blocking Fat Tree
Peak Power Consumption:13MW

FAQs

What is Summit?

Summit is Oak Ridge National Laboratory’s (ORNL’s) latest leadership-class supercomputer located at the ORNL Oak Ridge Leadership Computing Facility, a US Department of Energy Office of Science User Facility.

Dedicated to open science, Summit will drive scientific study and discovery as one of the world’s most powerful computers by enabling five to 10 times the computational performance of its predecessor, Titan.

A Summit node consists of two IBM Power9 CPUs, six NVIDIA V100 GPUs, NVLink for high-speed CPU-CPU and CPU-GPU communication, over half a terabyte of memory, and a large burst buffer for efficient I/O. Tying together about 4,600 of these nodes is a non-blocking fat-tree network of Mellanox EDR Infiniband, which provides high-performance internode communication and file system access.

With the V100, Summit enables unprecedented parallelism for traditional high-performance computing while also delivering to the scientific community one of the largest systems for artificial intelligence and deep learning. It’s a tool to solve some of the world’s most challenging problems, regardless of scientific domain, and to prepare computational science for future exascale systems.


What kind of data storage will be available on Summit?

In addition to other OLCF storage systems, such as the High-Performance Storage System for archival purposes, Summit will include a 250PB IBM Spectrum Scale file system. This parallel file system, named Alpine, will have performance of roughly 2.5TB/s of sequential I/O and 2.2TB/s of random I/O.


What is a Burst Buffer?

The burst buffer is an intermediate, high-speed layer of storage that is positioned between the application and the parallel file system (PFS), absorbing the bulk data produced by the application at a rate four to five times faster than the PFS, while seamlessly draining the data to the PFS in the background. Consequently, the burst buffer will be able to expedite I/O, allowing the application to return to performing computation sooner. The burst buffer is built from non-volatile memory devices that have several desirable properties such as high I/O throughput, low access latency, and higher reliability.


What is NVLINK?

GPUs in Titan are attached to each node by a traditional PCIe interface, which limits how fast the CPU memory system can be accessed. For increased performance, Summit makes use of NVIDIA’s NVLink interconnect for CPU-GPU and GPU-GPU communication. NVLink provides two links between every processor, each with a 25GB/s peak bandwidth in each direction. Supporting a peak bi-directional bandwidth of 100GB/s, these links are vital to the performance of accelerated applications on Summit.


How does the Unified Memory Feature
help?

The faster data movement that comes with NVLink, coupled with another feature known as Unified Memory, will simplify GPU accelerator programming. Unified Memory allows the programmer to treat the CPU and GPU memories as one block of memory. The programmer can operate on the data without worrying about whether it resides in the CPU’s or GPU’s memory.


What compilers will be available on Summit?

IBM XL, PGI, LLVM, GCC, NVIDIA CUDA Stack


What performance tools will be available on Summit?

MAP, Open|SpeedShop, TAU, HPCToolkit, VAMPIR/Score-P, Parallel Performance Toolkit, nvprof, gprof


What debugging tools will be available on Summit?

DDT, pdb, cuda-gdb, cuda-memcheck, valgrind, memcheck, helgrind, STAT


How much power will Summit consume? How does this compare to Titan?

Summit’s peak power consumption will be about 15 MW.


When will users get general access to Summit?

The plan of record is to provide early access to Summit for Early Science projects in late 2018 and to make Summit available to the OLCF User Programs starting in calendar year of 2019.


When will Titan be retired?

The plan of record is to keep Titan available for users for a period of time after Summit enters production.


What is Center for Accelerated Application Readiness (CAAR)?

The OLCF has created the Center for Accelerated Application Readiness, or CAAR, to help prepare codes for future generation systems. CAAR has established 13 partnership teams to prepare scientific applications for highly effective use on Summit. The partnership teams, consisting of the core developers of the application and staff from the OLCF, will receive support from the IBM/NVIDIA Center of Excellence at Oak Ridge National Laboratory and have access to multiple computational resources. For more information about CAAR, please visit olcf.ornl.gov/caar/.


What is CORAL?

CORAL is the collaboration between the two DOE Office of Science Leadership Computing Facility centers, Oak Ridge Leadership Computing Facility and Argonne Leadership Computing Facility, and the National Nuclear Security Administration Laboratory, Lawrence Livermore National Laboratory (LLNL) to procure leadership computer systems for their respective sites to support national security and scientific discovery. CORAL is an acronym for Collaboration of Oak Ridge, Argonne, and Livermore. For more information about CORAL, please visit the CORAL fact sheet.


What does ‘Leadership Computing’ mean?

“Leadership Computing” is a term that the DOE Office of Science uses to refer to the most powerful computing systems in the world. The Office of Science provides a portfolio of national high-performance computing facilities housing these supercomputers. These leadership computing facilities enable world-class research for significant advances in science.

The Oak Ridge Leadership Computing Facility (OLCF) was established at Oak Ridge National Laboratory in 2004 with the mission of accelerating scientific discovery and engineering progress by providing outstanding computing and data management resources to high-priority research and development projects.

Early Science Program

Note: this page is still live for archive purposes. The projects selected for the Summit Early Science program are listed here: Summit Early Science.

 

The OLCF is now accepting proposals for Early Science projects on Summit. The goals of the Early Science program are threefold. First, this will be an opportunity to realize early scientific achievements on what may well be the largest supercomputer for open science. Second, the Early Science projects will demonstrate the scalability and performance of applications ported to the Summit architecture. Third, the OLCF will benefit from the “hardening” of both the hardware and software environment using production-ready codes at scale in important mission-relevant scientific computationally challenging projects. The Early Science projects will have access to Summit from the time it has finished the acceptance testing period until the system is transitioning to the INCITE and ALCC user programs. While this will be a relatively short period of time, the capability of Summit is such that each project will have sufficient computational resources to carry out a challenging science campaign.


The Oak Ridge Leadership Computing Facility (OLCF) will deploy its next, pre-exascale system Summit in 2018. Summit is an IBM/NVIDIA system capable of over 200 petaflops delivered by ~4600 nodes with two IBM POWER9 processors and six NVIDIA V100 accelerators per node. Each node will have 512 GB of DDR4 memory, 96 GB of HBM2 memory connected to the GPUs, and 1600 GB of non-volatile memory, resulting in over 10 PB of total system memory. The node interconnect is a dual-rail EDR InfiniBand interconnect from Mellanox. Disk storage is provided by a 250 PB IBM GPFS filesystem.

FeatureTitanSummit
Application PerformanceBaseline5-10x Titan
Number of Nodes18,688~4,600
Node Performance1.4 TF> 40 TF
Memory per Node32 GB DDR3 + 6 GB GDDR5512 GB DDR4 + 96 GB HBM2
NV memory per Node01600 GB
Total System Memory710 TB>10 PB DDR4 + HBM2 + Non-volatile
System Interconnect (node injection bandwidth)Gemini (6.4 GB/s)Dual Rail EDR-IB (25 GB/s)
Interconnect Topology3D TorusNon-blocking Fat Tree
Processors1 AMD Opteron
1 NVIDIA Kepler
2 IBM POWER9
6 NVIDIA Volta
File System32 PB, 1 TB/s, Lustre®250 PB, 2.5 TB/s GPFS
Peak power consumption9 MW15 MW

The goals of the Early Science program are threefold. First, this will be an opportunity to realize early scientific achievements on what may well be the largest supercomputer for open science. Second, the Early Science projects will demonstrate the scalability and performance of applications ported to the Summit architecture. Third, the OLCF will benefit from the “hardening” of both the hardware and software environment using production-ready codes at scale in important mission-relevant scientific computationally challenging projects.

The Early Science projects will have access to Summit from the time it has finished the acceptance testing period until the system is transitioning to the INCITE and ALCC user programs. While this will be a relatively short period of time, the capability of Summit is such that each project will have sufficient computational resources to carry out a challenging science campaign.

Contact Information

Please direct any questions about this call for proposals by email to: [email protected]

Timeline for Summit Early Science

The anticipated timeline for Summit Early Science is as follows. The actual timeline will vary depending upon the realities and timing of the Summit deployment.

  • December 15, 2017: The call for Summit Early Science proposals opens;
  • December 31, 2017: Due date for letters of intent to respond to the Summit Early Science CFP;
  • Early 2018: As the first phase of Summit becomes available, proposing teams will be provided limited access for benchmarking runs required for preparation of the Summit Early Science proposal;
    • Proposals are encouraged to be submitted as soon as the necessary benchmarking results are available.
  • June 1, 2018: Due date for the proposals to the Summit Early Science CFP;
  • September 2018: Early Science projects selected;
  • October 1, 2018: Early Science projects have access to Summit. This date may change depending on the actual acceptance date of Summit;
  • January 1, 2019: Early Science project access to Summit reduces to 20% of Summit, as the INCITE user program transitions to Summit;
  • June 30, 2019: Early Science projects end;
  • Late 2019 (actual date TBD): Early Science project reports submission due;

Outcomes and Expectations of Early Science Projects

The Summit Early Science program is expected to deliver early scientific accomplishments in the form of peer-reviewed publications, reports, and highlights. This program is not for application development.

Support for Summit Early Science Projects

Each Summit Early Science project will be assigned an OLCF staff member as liaison. This liaison will devote a fraction of his or her time to collaborating with the Early Science project. Other technical experts will be involved as needed. These experts may include staff from the IBM/NVIDIA OLCF Center of Excellence or other ORNL research staff. Summit Early Science project teams will be invited to participate in Summit training events and workshops.

Summit Early Science Letter of Intent

A letter of intent to submit a Summit Early Science proposal is required, and due on December 31, 2017. This will enable the OLCF to arrange for reviewers, plan for the access to Summit needed to carry out the required benchmark calculations for the proposals, and plan for the allocation of Summit resources in the Early Science program period. These letters should include a list of members of the team and their affiliation, a paragraph with the science goal of the proposed project, a paragraph with a description of the scientific software to be used, including a brief statement on its current parallel scalability, accelerator performance, and the effort to make this application ready for Summit.

Early Science Proposal Submission

  • Submission deadline: June 1, 2018
    • Proposals will be accepted from the date of this call until the submission deadline. It is recommended to submit early, as proposals will be evaluated and awarded as they come in.
  • Proposals should be submitted by email to: [email protected]
  • Prepare your proposal using the instructions below
  • Submit as a single PDF document

Evaluation of Proposals

A team of internal ORNL staff and external science-domain experts will evaluate proposals on the following criteria.

  • Potential scientific impact of the proposed Early Science project
  • Demonstration of the need for Summit resources for the technical work
  • Demonstration of the computational readiness of the code, with respect scalability and accelerated performance
  • Expertise and commitment of the research team
  • Diversity of science domains and algorithms employed across the Early Science Program

Proposals will be evaluated as soon as possible after submission, with rolling acceptance into the Early Science program. Award decisions will be made by the OLCF-4 Project Team.

Proposal Instructions

Please create your proposal document with a project title, and the section headings noted below.

Section 1: Project Team

1a. Principal Investigator (PI) Information

  • Last Name, First Name, Title (Dr., Mr., Ms., etc.)
  • Institution
  • Street address
  • Email address
  • Funding source for the proposed work

1b. For each team member

  • Last Name, First Name, Title (Dr., Mr., Ms., etc.)
  • Institution
  • Street address
  • Email address
  • Funding source for the proposed work

Section 2: Project Description

2a. Executive Summary (1/2 page)

Provide an Executive Summary describing the scientific objectives and the impact of the proposed research. (1/2 page)

2b. Impact Statement (50 words)

Provide a two-sentence project summary that can be used to describe the impact of your project to the public.

2c. Scientific Campaign (2 pages)

Provide a technical description of the proposed project. Important aspects of this section are the scientific objectives, the scientific methodology, the relevance of the research, the need for resources provided by Summit, and the benefit of the project outcomes to the scientific community. A timeline and milestone table should be included.

Section 3: Computational Readiness (2-3 pages)

Provide a description of the computational readiness of the code that will be used in the science campaign. The overall objective is to demonstrate that the scientific campaign will utilize the Summit architecture efficiently, and that from a computational readiness perspective a proposal based on this code to the INCITE program would be competitive. For the Early Science proposals, computational readiness is, therefore, defined by a scalability and a performance metric as follows.

Scalability: Applications should demonstrate reduced time to solution (for strong scaling benchmarks) or time to solution divided by the number of nodes used (for weak scaling benchmarks) to 20% (i.e. 920 nodes) or more of the full Summit machine, N20. Timings should be obtained for several runs with node-counts below 920 nodes, one run at 920 nodes, and at least one run above 1000 nodes, in which both CPU’s and all GPU’s on the node are used.

Performance: Applications should demonstrate a performance improvement of a factor of two or better by using all six GPUs per node, compared to using both CPUs per node only, with jobs that runs on 20% (i.e. 920 nodes) of the full Summit machine. A description needs to be provided on the strategy used to optimize both cases.

The scalability and performance of the scientific code should be determined using runs that are representative of the calculations that will be carried out as part of the science campaign of the proposed project, and should be presented in a table with benchmark results, as well as in double logarithmic plots as follows.

The scalability and performance data need to be derived from benchmark calculations carried out on Summit. The OLCF will provide limited access to Summit during the time that the call is open.

Section 4: Resource Request (1 page)

Provide an estimate of the resources needed to meet the milestones described in section 2. Provide a detailed description of the number of runs, the number of nodes that will be used, and provide resource estimates based on the benchmark calculations described in section 3.

Section 5: Commitments

Please confirm that, should your proposal be awarded as a Summit Early Science project, you will commit to meeting the following requirements:

  1. Follow all OLCF policies and procedures
  2. Describe project progress during periodic conference calls with OLCF staff
  3. Submit a detailed project report at the end of the Early Science project period, to be received no later than July 31, 2019

CAAR Projects

In preparation for next-generation supercomputer Summit, the Oak Ridge Leadership Computing Facility (OLCF) selected 13 partnership projects into its Center for Accelerated Application Readiness (CAAR) program. A collaborative effort of application development teams and staff from the OLCF Scientific Computing group, CAAR is focused on redesigning, porting, and optimizing application codes for Summit’s hybrid CPU–GPU architecture. Through CAAR, codes teams gain access to early software development systems, leadership computing resources, and technical support from the IBM/NVIDIA Center of Excellence at Oak Ridge National Laboratory. The program culminates with each team’s scientific grand-challenge demonstration on Summit. The modeling and simulation applications selected for the CAAR program include:

Code: ACME
Science Domain: Climate
Title: Climate Research: Advancing Earth System Models
PI: David Bader, Lawrence Livermore National Laboratory

Code: DIRAC
Science Domain: Relativistic Quantum Chemistry
Title: CAAR Oak Ridge Proposal for getting the Relativistic Quantum Chemistry Program Package DIRAC ready for SUMMIT
PI: Lucas Visscher, Amsterdam Center for Multiscale Modeling /VU University Amsterdam

Code: FLASH
Science Domain: Astrophysics
Title: Using FLASH for Astrophysics Simulations at an Unprecedented Scale
PI: Bronson Messer, Oak Ridge National Laboratory

Code: GTC
Science Domain: Plasma Physics
Title: Particle Turbulence Simulations for Sustainable Fusion Reactions in ITER
PI: Zhihong Lin, University of California–Irvine

Code: HACC
Science Domain: Cosmology
Title: Cosmological Simulations for Large-scale Sky Surveys
PI: Salman Habib, Argonne National Laboratory

Code: LSDALTON
Science Domain: Quantum Chemistry
Title: Large-scale Coupled-cluster Calculations of Supramolecular Wires
PI: Poul Jørgensen, Aarhus University

Code: NAMD
Science Domain: Biophysics
Title: Molecular Machinery of the Brain
PI: Klaus Schulten, University of Illinois at Urbana-Champaign

Code: NUCCOR
Science Domain: Nuclear Physics
Title: Nuclear Structure and Nuclear Reactions
PI: Gaute Hagen, Oak Ridge National Laboratory

Code: NWCHEM
Science Domain: Computational Chemistry
Title: Developing Coupled Cluster Methodologies for GPUs
PI: Karol Kowalski, Pacific Northwest National Laboratory

Code: QMCPACK
Science Domain: Materials Sciences
Title: Materials Science Research for High-Temperature Superconductors
PI: Paul Kent, Oak Ridge National Laboratory

Code: RAPTOR
Science Domain: Engineering/Combustion
Title: Fluid Dynamics Research to Accelerate Combustion Science
PI: Joseph Oefelein, Sandia National Laboratories

Code: SPECFEM
Science Domain: Seismology
Title: Mapping the Earth’s Interior Using Big Data
PI: Jeroen Tromp, Princeton University

Code: XGC
Science Domain: Plasma Physics
Title: Multiphysics Magnetic Fusion Reactor Simulator, from Hot Core to Cold Wall
PI: C.S. Chang, Princeton Plasma Physics Laboratory, Princeton University

Announcing Summit

Summit is the next leap in leadership-class computing systems for open science. With Summit we will be able to address, with greater complexity and higher fidelity, questions concerning who we are, our place on earth, and in our universe.

Latest Summit Highlights

Filter

RTX

Summit Helps Forge Stronger Flights

Titanium alloys serve as cornerstone materials for the aerospace industry — stronger and lighter than steel, resistant to rust and corrosion and resilient past the melting points of most other metals. Companies such as RTX, formerly Raytheon Technologies, rely on these sturdy alloys to build such vital machinery as jet-engine turbine…
Matt LakinMatt LakinApril 30, 20248 min
X-ray bursts

Scientists Use Summit Supercomputer to Explore Exotic Stellar Phenomena

Understanding how a thermonuclear flame spreads across the surface of a neutron star — and what that spreading can tell us about the relationship between the neutron star’s mass and its radius — can also reveal a lot about the star’s composition. Neutron stars — the compact remnants of supernova…
Quinn BurkhartQuinn BurkhartMarch 15, 20246 min
NASA Mars simulation

Planning for a Smooth Landing on Mars

A U.S. mission to land astronauts on the surface of Mars will be unlike any other extraterrestrial landing ever undertaken by NASA. Although the space agency has successfully landed nine robotic missions on Mars since its first surface missions in 1976 with the Viking Project, safely bringing humans to Mars…
Coury TurczynCoury TurczynFebruary 29, 20249 min