Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Exascale computing”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 739 records · Page 41

Performance characterization of irregular I/O at the extreme scale

Here, this paper reports our experience with irregular I/O and describes lessons learned when running applications with such I/O on supercomputers at the extreme scale. Specifically, we study how irregularities in I/O patterns (i.e., irregular amount of data written per process at each I/O step) in scientific simulations can cause increasing I/O times and substantial loss in scalability. To this end, we quantify the impact of irregular I/O patterns on the I/O performance of scientific applications at the extreme scale by statistically modeling the irregular I/O behavior of two scientific applications: the Monte Carlo application QMCPack and the adaptive mesh refinement application ENZO. For our testing, we feed our model into I/O kernels of two well-known I/O data models (i.e., ADIOS and HDF) to measure the performance of the two applications’ I/O under different I/O settings. Empirically, we show how the growing data sizes and the irregular I/O patterns in these applications are both relevant factors impacting performance.

97 MATHEMATICS AND COMPUTING↗

FORGE: Pre-Training Open Foundation Models for Science

Large language models (LLMs) are poised to revolutionize the way we conduct scientific research. However, both model complexity and pre-training cost are impeding effective adoption for the wider science community. Identifying suitable scientific use cases, finding the optimal balance between model and data sizes, and scaling up model training are among the most pressing issues that need to be addressed. In this study, we provide practical solutions for building and using LLM-based foundation models targeting scientific research use cases. We present an end-to-end examination of the effectiveness of LLMs in scientific research, including their scaling behavior and computational requirements on Frontier, the first Exascale supercomputer. We have also developed for release to the scientific community a suite of open foundation models called FORGE with up to 26B parameters using 257B tokens from over 200M scientific articles, with performance either on par or superior to other state-of-the-art comparable models. We have demonstrated the use and effectiveness of FORGE on scientific downstream tasks. Our research establishes best practices that can be applied across various fields to take advantage of LLMs for scientific discovery.

Yin, Junqi↗

ornladios/ADIOS2

ADIOS 2: The Adaptable Input Output (I/O) System version 2 is an open-source framework that addresses scientific data management challenges, e.g. scalable parallel I/O, as we approach the exascale era in high-performance computing (HPC). ADIOS 2 bindings are available in C++, C, Fortran, Python and can be used on supercomputers, personal computers, and cloud systems running on Linux, macOS and Windows. ADIOS 2 has out-of-the-box support for MPI and serial environments.

ECP↗

Examination of a Methane/Diesel RCCI Engine Using Pele: Preprint

Multi-fuel, advanced injection strategies have become increasingly promising as a strategy to mitigate the emissions generated from internal combustion engines. By carefully controlling the combustion phasing in-cylinder, these new multi-pulse, multi-fuel injection strategies are able to burn in the low-temperature combustion regime where both NOx and soot are not readily produced, reducing the need for extensive exhaust gas recirculation systems. In this study, we examine a reactivity-controlled compression ignition (RCCI) strategy that uses an early pre-filled methane-air mixture with low turbulence background as the low-reactivity fuel and a direct injection of four discrete dodecane jets as a surrogate for the high-reactivity diesel fuel. We use the Pele software suite, a highly optimized, exascale-ready, adaptive mesh refinement codebase to perform high-resolution numerical simulations of a scaled down, single cylinder from the RCCI engine. Here, we resolve the ignition kernels down to micrometer scales and present several statistical quantities evaluating the development of the flow and detailing the onset of ignition and subsequent flame development. Particular attention is paid to the conditions surrounding the onset of the first ignition kernels and discussing what led to the development of those conditions.

CFD↗

Exascale Virtualized and Programmable Distributed Cyber Resource Control: Final Scientific Technical Report

Extreme-science drives the need for distributed exascale processing and communications that are carefully, yet flexibly, managed. Exponential growth of data for scientific simulations, experimental data, collaborative data analyses, remote visualization and GRID computing requirements of scientists in fields as diverse as high energy physics, climate change, genomics, fusion, synchrotron radiation, material science, medicine, and other scientific disciplines cannot be accommodated by simply applying existing transport protocols to faster pipes. Further, scientific challenges today demand diverse research teams, heightening the need for and increasing the complexity of collaboration. To address these issues within the network layer and physical layer, we have performed a number of research activities surrounding effective allocation and management of elastic optical network (EON) resources, particularly focusing on FlexGrid transponders. FlexGrid transponders support the opportunity to build Layer-1 connections at a wide range of bandwidths and to reconfigure them rapidly. The new flexibility supports complex new ways of using the physical layer that must be carefully managed and hidden from the scientist end-users. FlexGrid networks utilize flexible (or elastic) spectral bandwidths for each data link without using fixed wavelength grids. The flexibility in spectrum allocation brings many appealing features to network operations. Current networks are designed for the worst case impairments in transmission performance and the assigned spectrum is over-provisioned. In contrast, the FlexGrid networks can operate with the highest spectral efficiency and minimum bandwidth for the given traffic demand while meeting the minimum quality of transmission (QoT) requirement. Two primary focuses of our research are: (1) resource and spectrum allocation (RSA) for IP traffic over EONs, and (2) RSA for cross-domain optical networks. Previous work concentrates primarily on large file transfers within a single domain. Adding support for IP traffic changes the nature of the RSA problem: instead of choosing to accept or deny each request for network support, IP traffic is inherently elastic and thus lends itself to a bandwidth maximization formulation. We developed a number of algorithms that could be easily deployed within existing and new FlexGrid networks, leading to networks that better support scientific collaboration. Cross-domain RSA research is essential to support large-scale FlexGrid networks, since configuration information is generally not shared or coordinated across domains. The results presented here are in their early stages. They are technically feasible and practical, but still require coordination among organizations and equipment owners and a higher-layer framework for managing network requests.

42 ENGINEERING↗

A Comparison of Classical and Aggregation-Based Algebraic Multigrid Preconditioners for High-Fidelity Simulation of Wind Turbine Incompressible Flows

This paper presents a comparison of parallel strong scaling performance of classical and aggregation algebraic multigrid (AMG) preconditioners in the context of wind turbine simulations. Fluid motion is governed by the incompressible Navier--Stokes equations, discretized in space with control-volume finite elements and in time with an inexact projection scheme using an implicit integrator. A discontinuous-Galerkin sliding-mesh algorithm captures rotor motion. The momentum equations are solved with iterative Krylov methods, preconditioned by symmetric Gauss--Seidel (SGS) in Trilinos and $\ell_1$ SGS in hypre. The mass-continuity equation is solved with GMRES preconditioned by AMG and can account for the majority of simulation time. Reducing this continuity solve time is crucial. Wind turbine simulations present two unique challenges for AMG preconditioned solvers: the computational meshes include strongly anisotropic elements, and mesh motion requires matrix reinitialization and computation of preconditioners at each time step. Detailed timing profiles are presented and analyzed, and best practices are discussed for both classical and aggregation-based AMG. Results are presented for simulations of two different wind turbines with up to 6 billion grid points on two different computer architectures. For moving mesh problems that require linear-system reinitialization, the well-established strategy of amortizing preconditioner setup costs over a large number of time steps to reduce the solve time is no longer valid. Instead, results show that faster time to solution is achieved by reducing preconditioner setup costs at the expense of linear-system solve costs. Standard smoothed aggregation with Chebyshev relaxation was found to perform poorly when compared with classical AMG in terms of solve time and robustness. However, plain aggregation was comparable to classical AMG.

97 MATHEMATICS AND COMPUTING↗

Scalable Contour Tree Computation by Data Parallel Peak Pruning

As data sets grow to exascale, automated data analysis and visualization are increasingly important, to intermediate human understanding and to reduce demands on disk storage via in situ analysis. Trends in architecture of high performance computing systems necessitate analysis algorithms to make effective use of combinations of massively multicore and distributed systems. One of the principal analytic tools is the contour tree, which analyses relationships between contours to identify features of more than local importance. Unfortunately, the predominant algorithms for computing the contour tree are explicitly serial, and founded on serial metaphors, which has limited the scalability of this form of analysis. While there is some work on distributed contour tree computation, and separately on hybrid GPU-CPU computation, there is no efficient algorithm with strong formal guarantees on performance allied with fast practical performance. Furthermore, we report the first shared SMP algorithm for fully parallel contour tree computation, with formal guarantees of O(lg V lg t) parallel steps and O(V lg V) work for data with V samples and t contour tree supernodes, and implementations with more than 30× parallel speed up on both CPU using TBB and GPU using Thrust and up 70× speed up compared to the serial sweep and merge algorithm.

97 MATHEMATICS AND COMPUTING↗

Optimizing Distributed Training on Frontier for Large Language Models

Large language models (LLMs) have demonstrated remarkable success as foundational models, benefiting various downstream applications through fine-tuning. Loss scaling studies have demonstrated the superior performance of larger LLMs compared to their smaller counterparts. Nevertheless, training LLMs with billions of parameters poses significant challenges and requires considerable computational resources. For example, training a one trillion parameter GPT-style model on 20 trillion tokens requires a staggering 120 million exaflops. This research explores efficient distributed training strategies to extract this computation from Frontier, the world's first exascale supercomputer. We enable and investigate various model and data parallel training techniques, such as tensor parallelism, pipeline parallelism, and sharded data parallelism, to facilitate training a trillion-parameter model on Frontier. We empirically assess these techniques and their associated parameters to determine their impact on memory footprint, communication latency, and GPU's computational efficiency. We analyze the complex interplay among these techniques and find a strategy to combine them to achieve high throughput through hyperparameter tuning. We have identified efficient strategies for training large LLMs of varying sizes through empirical analysis and hyperparameter tuning. For 22 Billion, 175 Billion, and 1 Trillion parameters, we achieved GPU throughputs of 38.38%, 36.14%, and 31.96%, respectively. For the training of the 175 Billion parameter model and the 1 Trillion parameter model, we achieved 100% weak scaling efficiency on 1024 and 3072 Mi250X GPUs, respectively. We also achieved strong scaling efficiencies of 89% and 87% for these two models. We trained these models only tens of iterations instead of training till completion.

Yin, Junqi↗

Analytical Cost Metrics : Days of Future Past

As we move towards the exascale era, the new architectures must be capable of running the massive computational problems efficiently. Scientists and researchers are continuously investing in tuning the performance of extreme-scale computational problems. These problems arise in almost all areas of computing, ranging from big data analytics, artificial intelligence, search, machine learning, virtual/augmented reality, computer vision, image/signal processing to computational science and bioinformatics. With Moore’s law driving the evolution of hardware platforms towards exascale, the dominant performance metric (time efficiency) has now expanded to also incorporate power/energy efficiency. Therefore the major challenge that we face in computing systems research is: “how to solve massive-scale computational problems in the most time/power/energy efficient manner?”

97 MATHEMATICS AND COMPUTING↗

Quantifying Scheduling Challenges for Exascale System Software

The move towards high-performance computing (HPC) ap- plications comprised of coupled codes and the need to dra- matically reduce data movement is leading to a reexami- nation of time-sharing vs. space-sharing in HPC systems. In this paper, we discuss and begin to quantify the perfor- mance impact of a move away from strict space-sharing of nodes for HPC applications. Specifically, we examine the po- tential performance cost of time-sharing nodes between ap- plication components, we determine whether a simple coor- dinated scheduling mechanism can address these problems, and we research how suitable simple constraint-based opti- mization techniques are for solving scheduling challenges in this regime. Our results demonstrate that current general- purpose HPC system software scheduling and resource al- location systems are subject to significant performance de- ciencies which we quantify for six representative applica- tions. Based on these results, we discuss areas in which ad- ditional research is needed to meet the scheduling challenges of next-generation HPC systems.

C.1.2 [Multiprocessors]: Parallel Processors↗

A survey of CPU-GPU heterogeneous computing techniques

As both CPU and GPU become employed in a wide range of applications, it has been acknowledged that both of these processing units (PUs) have their unique features and strengths and hence, CPU-GPU collaboration is inevitable to achieve high-performance computing. This has motivated significant amount of research on heterogeneous computing techniques, along with the design of CPU-GPU fused chips and petascale heterogeneous supercomputers. In this paper, we survey heterogeneous computing techniques (HCTs) such as workload-partitioning which enable utilizing both CPU and GPU to improve performance and/or energy efficiency. We review heterogeneous computing approaches at runtime, algorithm, programming, compiler and application level. Further, we review both discrete and fused CPU-GPU systems; and discuss benchmark suites designed for evaluating heterogeneous computing systems (HCSs). Furthermore, we believe that this paper will provide insights into working and scope of applications of HCTs to researchers and motivate them to further harness the computational powers of CPUs and GPUs to achieve the goal of exascale performance.

97 MATHEMATICS AND COMPUTING↗

Toward Exascale: Overview of Large Eddy Simulations and Direct Numerical Simulations of Nuclear Reactor Flows with the Spectral Element Method in Nek5000

At the beginning of the last decade, Petascale supercomputers (i.e., computers capable of more than 1 petaFLOP) emerged. Now, at the dawn of exascale supercomputing, we provide a review of recent landmark simulations of portions of reactor components with turbulence-resolving techniques that this computational power has made possible. In fact, these simulations have provided invaluable insight into flow dynamics, which is difficult or often impossible to obtain with experiments alone. We focus on simulations performed with the spectral element method, as this method has emerged as a powerful tool to deliver massively parallel calculations at high fidelity by using large eddy simulation or direct numerical simulation. We also limit this paper to constant-property incompressible flow of a Newtonian fluid in the absence of other body or external forces, although the method is by no means limited to this class of flows. We briefly review the fundamentals of the method and the reasons it is compelling for the simulation of nuclear engineering flows. We review in detail a series of Petascale simulations, including the simulations of helical coil steam generators, fuel assemblies, and pebble beds. Even with Petascale computing, however, limitations for nuclear modeling and simulation tools remain. In particular, the size and scope of turbulence-resolving simulations are still limited by computing power and resolution requirements, which scale with the Reynolds number. In the final part of this paper, we discuss the future of the field, including recent advancements in emerging architectures such as GPUbased supercomputers, which are expected to power the next generation of high-performance computers.

computational fluid dynamics↗

A Pervasive Parallel Processing Framework for Data Visualization and Analysis at Extreme Scale

The evolution of the computing world from teraflop to petaflop has been relatively effortless, with several of the existing programming models scaling effectively to the petascale. The migration to exascale, however, poses considerable challenges. All industry trends infer that the exascale machine will be built using processors containing hundreds to thousands of cores per chip. It can be inferred that efficient concurrency on exascale machines requires a massive amount of concurrent threads, each performing many operations on a localized piece of data. Currently, visualization libraries and applications are based off what is known as the visualization pipeline. In the pipeline model, algorithms are encapsulated as filters with inputs and outputs. These filters are connected by setting the output of one component to the input of another. Parallelism in the visualization pipeline is achieved by replicating the pipeline for each processing thread. This works well for today’s distributed memory parallel computers but cannot be sustained when operating on processors with thousands of cores. Our project investigates a new visualization framework designed to exhibit the pervasive parallelism necessary for extreme scale machines. Our framework achieves this by defining algorithms in terms of worklets, which are localized stateless operations. Worklets are atomic operations that execute when invoked unlike filters, which execute when a pipeline request occurs. The worklet design allows execution on a massive amount of lightweight threads with minimal overhead. Only with such fine-grained parallelism can we hope to fill the billions of threads we expect will be necessary for efficient computation on an exascale machine.

97 MATHEMATICS AND COMPUTING↗

Operating System Support for Fine-Grained Task Migration

With the end of clock scaling and the limited power budget available (20-30MW), future supercomputers will meet exascale performance primarily through a higher level of parallelism. Current operating (OS) and runtime (RT) systems are designed for the classical SMP model and based on the static and coarse-grained process/thread paradigm. They do not provide the required level of flexibility, especially within a single compute node, to meet the requirements imposed by exascale systems in terms of power/energy efficiency, resilience, managing concurrency and performance portability. The high level of concurrency poses new challenges specific to exascale systems that need to be addressed by novel solutions. In particular, requiring the user to manage billions of concurrent threads could easily result in poor data locality, clogged interconnection networks, unchecked propagation of soft errors, and lack of control over power/energy consumption. Equally important, managing such level of concurrency interferes with the user’s focus on the application and the algorithm. We envision that the computation will be encapsulated into fine-grained tasks that can be isolated and protected from the other tasks running in the system. Whenever a task needs to work on some data that is not stored on the local node, the OS/RT allows the task to be migrated to the node that owns the data. Each task is associated with a contained state (set or architectural registers, stack frame, running node, etc.) that describes the progress of the task and that should be moved together with the task’s code. We believe system support for task migration is a fundamental function that can ease the job of tackling several of the exascale challenges.

Gioiosa, Roberto↗

Making Uintah Performance Portable for Department of Energy Exascale Testbeds

To help ease ports to forthcoming Department of Energy (DOE) exascale systems, testbeds have been made available to select users. These testbeds are helpful for preparing codes to run on the same hardware and similar software as in their respective exascale systems. This paper describes how the Uintah Computational Framework, an open-source asynchronous many-task (AMT) runtime system, has been modified to be performance portable across the DOE Crusher, DOE Polaris, and DOE Sunspot testbeds in preparation for portable simulations across the exascale DOE Frontier and DOE Aurora systems. The Crusher, Polaris, and Sunspot testbeds feature the AMD MI250X, NVIDIA A100, and Intel PVC GPUs, respectively. This performance portability has been made possible by extending Uintah’s intermediate portability layer [18] to additionally support the Kokkos::HIP, Kokkos::OpenMPTarget, and Kokkos::SYCL back-ends. This paper also describes notable updates to Uintah’s support for Kokkos, which were required to make this extension possible. Results are shown for a challenging radiative heat transfer calculation, central to the University of Utah’s predictive boiler simulations. These results demonstrate single-source portability across AMD-, NVIDIA-, and Intel-based GPUs using various Kokkos back-ends.

Holmen, John↗

Reducing Communication in Algebraic Multigrid Using Additive Variants

Algebraic multigrid (AMG) has proven to be an effective scalable solver on many high performance computers. However, its increasing communication complexity on coarser levels has shown to seriously impact its performance on computers with high communication cost. Moreover, additive AMG variants provide not only increased parallelism as well as decreased numbers of messages per cycle but also generally exhibit slower convergence. Here we present various new additive variants with convergence rates that are significantly improved compared to the classical additive algebraic multigrid method and investigate their potential for decreased communication, and improved communication-computation overlap, features that are essential for good performance on future exascale architectures.

97 MATHEMATICS AND COMPUTING↗

Report of experiments and evidence for ASC L2 milestone 4467: demonstration of a legacy application's path to exascale

This report documents thirteen of Sandia's contributions to the Computational Systems and Software Environment (CSSE) within the Advanced Simulation and Computing (ASC) program between fiscal years 2009 and 2012. It describes their impact on ASC applications. Most contributions are implemented in lower software levels allowing for application improvement without source code changes. Improvements are identified in such areas as reduced run time, characterizing power usage, and Input/Output (I/O). Other experiments are more forward looking, demonstrating potential bottlenecks using mini-application versions of the legacy codes and simulating their network activity on Exascale-class hardware. The purpose of this report is to prove that the team has completed milestone 4467-Demonstration of a Legacy Application's Path to Exascale. Cielo is expected to be the last capability system on which existing ASC codes can run without significant modifications. This assertion will be tested to determine where the breaking point is for an existing highly scalable application. The goal is to stretch the performance boundaries of the application by applying recent CSSE RD in areas such as resilience, power, I/O, visualization services, SMARTMAP, lightweight LWKs, virtualization, simulation, and feedback loops. Dedicated system time reservations and/or CCC allocations will be used to quantify the impact of system-level changes to extend the life and performance of the ASC code base. Finally, a simulation of anticipated exascale-class hardware will be performed using SST to supplement the calculations. Determine where the breaking point is for an existing highly scalable application: Chapter 15 presented the CSSE work that sought to identify the breaking point in two ASC legacy applications-Charon and CTH. Their mini-app versions were also employed to complete the task. There is no single breaking point as more than one issue was found with the two codes. The results were that applications can expect to encounter performance issues related to the computing environment, system software, and algorithms. Careful profiling of runtime performance will be needed to identify the source of an issue, in strong combination with knowledge of system software and application source code.

99 GENERAL AND MISCELLANEOUS↗

Pre-exascale accelerated application development: The ORNL Summit experience

High-performance computing (HPC) increasingly relies on heterogeneous architectures to achieve higher performance. In the Oak Ridge Leadership Facility (OLCF), Oak Ridge, TN, USA, this trend continues as its latest supercomputer, Summit, entered production in early 2019. The combination of IBM POWER9 CPU and NVIDIA V100 GPU, along with a fast NVLink2 interconnect and other latest technologies, pushes system performance to a new height and breaks the exascale barrier by certain measures. Due to Summit's powerful GPUs and much higher GPU–CPU ratio, offloading to accelerators becomes a requirement for any application, which intends to effectively use the system. To facilitate navigating a complex landscape of competing heterogeneous architectures, a collection of applications from a wide spectrum of scientific domains is selected for early adoption on Summit. In this article, the experience and lessons learned are summarized, in the hope of providing useful guidance to address new programming challenges, such as scalability, performance portability, and software maintainability, for future application development efforts on heterogeneous HPC systems.

97 MATHEMATICS AND COMPUTING↗