Search NASA⌕ Search

SEARCH · Search NASA

Results for “GPU accelerators”

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 307 records · Page 17

Preparing an Incompressible-Flow Fluid Dynamics Code for Exascale-Class Wind Energy Simulations

The U.S. Department of Energy has identified exascale-class wind farm simulation as critical to wind energy scientific discovery. A primary objective of the ExaWind project is to build high-performance, predictive computational fluid dynamics (CFD) tools that satisfy these modeling needs. GPU accelerators will serve as the computational thoroughbreds of next-generation, exascale-class supercomputers. Here, we report on our efforts in preparing the ExaWind unstructured mesh solver, Nalu-Wind, for exascale-class machines. For computing at this scale, a simple port of the incompressible-flow algorithms to GPUs is insufficient. To achieve high performance, one needs novel algorithms that are application aware, memory efficient, and optimized for the latest-generation GPU devices. The result of our efforts are unstructured-mesh simulations of wind turbines that can effectively leverage thousands of GPUs. In particular, we demonstrate a first-of-its-kind, incompressible-flow simulation using Algebraic Multigrid solvers that strong scales to more than 4000 GPUs on the Summit supercomputer.

algebraic multigrid↗

Evolution of the SLATE linear algebra library

SLATE (Software for Linear Algebra Targeting Exascale) is a distributed, dense linear algebra library targeting both CPU-only and GPU-accelerated systems, developed over the course of the Exascale Computing Project (ECP). While it began with several documents setting out its initial design, significant design changes occurred throughout its development. In some cases, these were anticipated: an early version used a simple consistency flag that was later replaced with a full-featured consistency protocol. In other cases, performance limitations and software and hardware changes prompted a redesign. Sequential communication tasks were parallelized; host-to-host MPI calls were replaced with GPU device-to-device MPI calls; more advanced algorithms such as Communication Avoiding LU and the Random Butterfly Transform (RBT) were introduced. Early choices that turned out to be cumbersome, error prone, or inflexible have been replaced with simpler, more intuitive, or more flexible designs. Applications have been a driving force, prompting a lighter weight queue class, nonuniform tile sizes, and more flexible MPI process grids. Of paramount importance has been building a portable library that works across several different GPU architectures – AMD, Intel, and NVIDIA – while keeping a clean and maintainable codebase. Here we explore the evolving design choices and their effects, both in terms of performance and software sustainability.

Gates, Mark↗

Farpoint: A High-resolution Cosmology Simulation at the Gigaparsec Scale

Abstract In this paper we introduce the Farpoint simulation, the latest member of the Hardware/Hybrid Accelerated Cosmology Code (HACC) gravity-only simulation family. The domain covers a volume of (1000 h −1 Mpc) 3 and evolves close to two trillion particles, corresponding to a mass resolution of m p ∼ 4.6 × 10 7 h −1 M ⊙ . These specifications enable comprehensive investigations of the galaxy–halo connection, capturing halos down to small masses. Further, the large volume resolves scales typical of modern surveys with good statistical coverage of high-mass halos. The simulation was carried out on the GPU-accelerated system Summit, one of the fastest supercomputers currently available. We provide specifics about the Farpoint run and present an initial set of results. The high mass resolution facilitates precise measurements of important global statistics, such as the halo concentration–mass relation and the correlation function down to small scales. Selected subsets of the simulation data products are publicly available via the HACC Simulation Data Portal.

79 ASTRONOMY AND ASTROPHYSICS↗

PYSEQM

PYSEQM is a package for performing semi-empirical quantum mechanical (SEQM) simulations on molecular systems utilizing PyTorch. SEQM simulations determine molecular properties (energy, electron density, dipole moment, ect.) by solving an approximate Schrödinger equation for the motions of electrons in a molecule. The use of PyTorch provides three specific advantages. First, it allows the calculations to be offloaded to GPU accelerators, giving an order of magnitude increase in speed. Second, back propagation is used to get atomic forces (derivative of the total energy with respect to atomic position) at the same computational cost as the energy calculation itself. Finally, the use of PyTorch makes for a natural interface to modern machine learning methods, which can be used to adjust the semi-empirical parameters build into SEQM methods. Additionally, PYSEQM implements various other optimizations for performing quantum mechanics based molecular dynamics, including SP2 for rapid GPU based solution of the self-consistent field algorithm, and the extended Lagrangian method for rapid QM-MD.

Nebgen, Benjamin↗

Accelerating Multigrid-based Hierarchical Scientific Data Refactoring on GPUs

Rapid growth in scientific data and a widening gap between computational speed and I/O bandwidth make it increasingly infeasible to store and share all data produced by scientific simulations. Instead, we need methods for reducing data volumes: ideally, methods that can scale data volumes adaptively so as to enable negotiation of performance and fidelity tradeoffs in different situations. Multigrid-based hierarchical data representations hold promise as a solution to this problem, allowing for flexible conversion between different fidelities so that, for example, data can be created at high fidelity and then transferred or stored at lower fidelity via logically simple and mathematically sound operations. However, the effective use of such representations has been hindered until now by the relatively high costs of creating, accessing, reducing, and otherwise operating on such representations. We describe here highly optimized data refactoring kernels for GPU accelerators that enable efficient creation and manipulation of data in multigrid-based hierarchical forms. We demonstrate that our optimized design can achieve up to 250 TB/s aggregated data refactoring throughput—83% of theoretical peak—on 1024 nodes of the Summit supercomputer. We showcase our optimized design by applying it to a large-scale scientific visualization workflow and the MGARD lossy compression software.

Data refactoring↗

A Length Adaptive Algorithm-Hardware Co-design of Transformer on FPGA Through Sparse Attention and Dynamic Pipelining

Transformers are considered one of the most important deep learning models since 2018, in part because it establishes state-of-the-art (SOTA) records and could potentially replace existing Deep Neural Networks (DNNs). Despite the remarkable triumphs, the prolonged turnaround time of Transformer models is a widely recognized roadblock. The variety of sequence lengths imposes additional computing overhead where inputs need to be zero-padded to the maximum sentence length in the batch to accommodate the parallel computing platforms. This paper targets the field-programmable gate array (FPGA) and proposes a coherent sequence length adaptive algorithm–hardware co-design for Transformer acceleration. Particularly, we develop a hardware-friendly sparse attention operator and a length-aware hardware resource scheduling algorithm. The proposed sparse attention operator brings the complexity of attention-based models down to linear complexity and alleviates the off-chip memory traffic. The proposed length-aware resource hardware scheduling algorithm dynamically allocates the hardware resources to fill up the pipeline slots and eliminates bubbles for NLP tasks. Experiments show that our design has very small accuracy loss and has 80.2 × and 2.6 × speedup compared to CPU and GPU implementation, and 4 × higher energy efficiency than state-of-the-art GPU accelerator optimized via CUBLAS GEMM.

Peng, Hongwu↗

Nek5000/RS Performance on Advanced GPU Architectures

We demonstrate NekRS performance results on various advanced GPU architectures. NekRS is a GPU-accelerated version of Nek5000 that is targeting high performance on forthcoming exascale platforms. It is being developed in DOE’s Center of Efficient Exascale Discretizations (CEED), which is one of the co-design centers under the Exascale Computing Project (ECP). In this report, we consider Frontier, Crusher, Spock, Polaris, Perlmutter, ThetaGPU and Summit. Simulations are performed using ExaSMR’s 17x17 rod-bundle geometries with different problem sizes. The report focuses on strong scaling performance and analysis. Many of the results shown in this report are the outcome from participation in the ALCF GPU Hackathon on Polaris, which was held on 7/19/22, and 7/26–7/28/22. Members of the Nek5000/RS team included Misun Min, Yu-Hsiang Lan, Paul Fischer and Thilina Rathnayake. Mentors were Kris Rowe (ALCF) and Peng Wang (NVIDIA). Also presented in this report are results on Frontier, which were obtained in collaboration with John Holmen at OLCF.

97 MATHEMATICS AND COMPUTING↗

Early experiences evaluating the HPE/Cray ecosystem for AMD GPUs

Summary The Oak Ridge Leadership Computing Facility (OLCF) has a long history of supporting and promoting GPU‐accelerated computing starting with the deployment of the Titan supercomputer in 2021 and continuing with the Summit supercomputer which has a theoretical peak performance of approximately 200 petaflops. Because the majority of Summit's computational power comes from its 27,972 GPUs, users must port their applications to one of the supported programming models in order to make efficient use of the system. To prepare the transition to Frontier, the OLCF's exascale supercomputer, users will need to adapt to an entirely new ecosystem which will include new hardware and software technologies. First, users will need to familiarize themselves with the AMD Radeon GPU architecture. Furthermore, users who have been previously relying on CUDA will need to transition to the Heterogeneous‐Computing Interface for Portability (HIP) or one of the other supported programming models (e.g., OpenMP, OpenACC). In this work, we describe our initial experiences and lessons learned in porting three applications or proxy apps currently running on Summit to the HPE/Cray ecosystem to leverage the compute power from AMD GPUs: minisweep, GenASiS, and Sparkler. Each one is representative of current production workloads utilized at the OLCF, different programming languages, and different programming models.

Melesse Vergara, Verónica G.↗

Optimizing Error-Bounded Lossy Compression for Scientific Data on GPUs

Error-bounded lossy compression is a critical technique for significantly reducing scientific data volumes. With ever-emerging heterogeneous high-performance computing (HPC) architecture, GPU-accelerated error-bounded compressors (such as CUSZ and cuZFP) have been developed. However, they suffer from either low performance or low compression ratios. To this end, we propose CUSZ+ to target both high compression ratios and throughputs. We identify that data sparsity and data smoothness are key factors for high compression throughputs. Our key contributions in this work are fourfold: (1) We propose an efficient compression workflow to adaptively perform run-length encoding and/or variable-length encoding. (2) We derive Lorenzo reconstruction in decompression as multidimensional partial-sum computation and propose a fine-grained Lorenzo reconstruction algorithm for GPU architectures. (3) We carefully optimize each of CUSZ kernels by leveraging state-of-the-art CUDA parallel primitives. (4) We evaluate CUSZ+ using seven real-world HPC application datasets on V100 and A100 GPUs. Experiments show CUSZ+ improves the compression throughputs and ratios by up to 18.4x and 5.3x, respectively, over CUSZ on the tested datasets.

Tian, Jiannan↗

Accelerating Multigrid-based Hierarchical Scientific Data Refactoring on GPUs

Rapid growth in scientific data and a widening gap between computational speed and I/O bandwidth make it increasingly infeasible to store and share all data produced by scientific simulations. Instead, we need methods for reducing data volumes: ideally, methods that can scale data volumes adaptively so as to enable negotiation of performance and fidelity tradeoffs in different situations. Multigrid-based hierarchical data representations hold promise as a solution to this problem, allowing for flexible conversion between different fidelities so that, for example, data can be created at high fidelity and then transferred or stored at lower fidelity via logically simple and mathematically sound operations. However, the effective use of such representations has been hindered until now by the relatively high costs of creating, accessing, reducing, and otherwise operating on such representations. We describe here highly optimized data refactoring kernels for GPU accelerators that enable efficient creation and manipulation of data in multigrid-based hierarchical forms. We demonstrate that our optimized design can achieve up to 250 TB/s aggregated data refactoring throughput—83% of theoretical peak—on 1024 nodes of the Summit supercomputer. We showcase our optimized design by applying it to a large-scale scientific visualization workflow and the MGARD lossy compression software.

Chen, Jieyang↗

Simulating Hydrodynamics in Cosmology with CRK-HACC

Abstract We introduce CRK-HACC, an extension of the Hardware/Hybrid Accelerated Cosmology Code (HACC), to resolve gas hydrodynamics in large-scale structure formation simulations of the universe. The new framework couples the HACC gravitational N -body solver with a modern smoothed-particle hydrodynamics (SPH) approach called conservative reproducing kernel SPH (CRKSPH). CRKSPH utilizes smoothing functions that exactly interpolate linear fields while manifestly preserving conservation laws (momentum, mass, and energy). The CRKSPH method has been incorporated to accurately model baryonic effects in cosmology simulations—an important addition targeting the generation of precise synthetic sky predictions for upcoming observational surveys. CRK-HACC inherits the codesign strategies of the HACC solver and is built to run on modern GPU-accelerated supercomputers. In this work, we summarize the primary solver components and present a number of standard validation tests to demonstrate code accuracy, including idealized hydrodynamic and cosmological setups, as well as self-similarity measurements.

79 ASTRONOMY AND ASTROPHYSICS↗

SQMBox: Interfacing a semiempirical integral library to modular ab initio electronic structure enables new semiempirical methods

Ab initio and semiempirical electronic structure methods are usually implemented in separate software packages or use entirely different code paths. As a result, it can be time-consuming to transfer an established ab initio electronic structure scheme to a semiempirical Hamiltonian. Here we present an approach to unify ab initio and semiempirical electronic structure code paths based on a separation of the wavefunction ansatz and the needed matrix representations of operators. With this separation, the Hamiltonian can refer to either an ab initio or semiempirical treatment of the resulting integrals. We built a semiempirical integral library and interfaced it to the GPU-accelerated electronic structure code TeraChem. Equivalency between ab initio and semiempirical tight-binding Hamiltonian terms is assigned according to their dependence on the one-electron density matrix. The new library provides semiempirical equivalents of the Hamiltonian matrix and gradient intermediates, corresponding to those provided by the ab initio integral library. This enables the straightforward combination of semiempirical Hamiltonians with the full pre-existing ground and excited state functionality of the ab initio electronic structure code. We demonstrate the capability of this approach by combining the extended tight-binding method GFN1-xTB with both spin-restricted ensemble-referenced Kohn–Sham and complete active space methods. We also present a highly efficient GPU implementation of the semiempirical Mulliken-approximated Fock exchange. The additional computational cost for this term becomes negligible even on consumer-grade GPUs, enabling Mulliken-approximated exchange in tight-binding methods for essentially no additional cost.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Bolt: A Fast Solver for Kinetic Theories Using a High-Resolution Constrained Transport Scheme [Slides]

Our understanding of collisionless and semi-collisional plasmas in the nonlinear regime is limited by the expense of computing solutions numerically. Bolt is a fast, GPU-accelerated code for rapidly computing such solutions with accurate transport and an approximate collision operator. Such calculations are relevant to both problems in astrophysics, such as heat conduction and magnetic reconnection in accretion disks around black holes, and also to programmatic interests at LANL. The bolt code paper, demonstrating accuracy via a suite of test problems calculated on kodiak, is currently in preparation.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Parallel quantum computing simulations via quantum accelerator platform virtualization

Quantum circuit execution is a central task in quantum computation. Due to inherent quantum-mechanical constraints, quantum computing workflows often involve a considerable number of independent measurements over a large set of slightly different quantum circuits. Here we discuss a simple model for parallelizing such quantum circuit executions that is based on introducing a large array of virtual quantum processing units (mapped to HPC nodes in our case) as a parallel quantum computing platform. Implemented within the XACC framework, the model can readily take advantage of its backend-agnostic features, enabling parallel quantum computing/simulation over any target backend supported by XACC. We illustrate the performance of this approach by demonstrating strong scaling in two pertinent domain science problems, namely in computing the gradients for the multi-contracted variational quantum eigensolver and in data-driven quantum circuit learning, where we vary the number of qubits and the number of circuit layers. Here, the latter simulation leverages the cuQuantum library to run efficiently on GPU-accelerated HPC platforms.

97 MATHEMATICS AND COMPUTING↗

2D numerical simulation of unsteady flows for large scale floods prediction in real time

The challenge of finding a compromise between computational time and level of accuracy and robustness has traditionally expanded the use simplified models rather than full two-dimensional (2D) models for flood simulation. This work presents a GPU accelerated 2D shallow water model for the simulation of flood events in real time. In particular, an explicit first-order finite volume scheme is detailed to control the numerical instabilities that are likely to appear when used in complex topography. The model is first validated with the benchmark test case of the Toce River (Italy) and numerical fixes are demonstrated to be necessary. The model is next applied to reproduce real events in a reach of the Ebro River (Spain) in order to compare simulation results with field data. The second case deals with a large domain (744 km 2 ) and long flood duration (up to 20 days) allowing an analysis of the performance and speed-up achieved by different GPU devices. We report the high values of fit between observed and simulated results as well as the computational times achieved are encouraging to propose the use of the model as forecasting system.

54 ENVIRONMENTAL SCIENCES↗

Simulating Electron Dynamics of Complex Molecules with Time-Dependent Complete Active Space Configuration Interaction

In this paper, time-dependent electronic structure methods are growing in popularity as tools for modeling ultrafast and/or nonlinear processes, for computing spectra, and as the electronic structure component of mean-field molecular dynamics simulations. Time-dependent configuration interaction (TD-CI) offers several advantages over the widely used real-time time-dependent density functional theory: namely, that it correctly models Rabi oscillations; it offers a spin-pure description of open-shell systems; and a hierarchy of TD-CI methods can be defined that systematically approach the exact solution of the time-dependent Schrodinger equation (TDSE). In this work, we present a novel TD-CI approach that extends TD-CI to large complete active-space configuration expansions. Such extension is enabled by use of a direct configuration interaction approach that eliminates the need to explicitly build, store, or diagonalize the Hamiltonian matrix. Graphics processing unit (GPU) acceleration enables fast solution of the TDSE even for large active spaces—up to 12 electrons in 12 orbitals (853776 determinants) in this work. A symplectic split operator propagator yields long-time norm conservation. We demonstrate the applicability of our approach by computing the response of a large molecule with a strongly correlated ground state, decacene (C 42 H 24 ), to various pulses (δ-function, transform limited, chirped). Our simulations predict that chirped pulses can be used to induce dipole-forbidden transitions. Simulations of decacene using the 6-31G(d) basis set and a 12 electrons/12 orbitals active space took 20.1 h to propagate for 100 fs with a 1 attosecond time step on a single NVIDIA K40 GPU. Convergence with respect to time step is found to depend on the property being computed and the chosen active space.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Accelerating the Computation of Detailed Chemical Reaction Kinetics for Simulating Combustion of Complex Fuels

Combustion of hydrocarbon fuels has been a very challenging scientific and engineering problem due to the complexity of turbulent flows and hydrocarbon reaction kinetics. There is an urgent need to develop an efficient modeling capability to accurately predict the combustion of complex fuels. Detailed chemical kinetic models for the surrogates of fuels such as gasoline, diesel and JP-8 consist of thousands of chemical species and Arrhenius reaction steps. Oxygenated fuels such as bio-fuels and heavier hydrocarbons, such as from newer fossil fuel sources, are expected to have a much more complex chemistry requiring increasingly larger chemical kinetic models. Such models are beyond current computational capability, except for homogeneous or partially stirred reactor type calculations. The advent of highly parallel multi-core processors and graphical processing units (GPUs) promises a steep increase in computational performance in the coming years. This paper will present a software framework that translates the detailed chemical kinetic models to high- performance code targeted for GPU accelerators.

02 PETROLEUM↗

Accelerating the Computation of Detailed Chemical Reaction Kinetics for Simulating Combustion of Complex Fuels

Combustion of hydrocarbon fuels has been a very challenging scientific and engineering problem due to the complexity of turbulent flows and hydrocarbon reaction kinetics. There is an urgent need to develop an efficient modeling capability to accurately predict the combustion of complex fuels. Detailed chemical kinetic models for the surrogates of fuels such as gasoline, diesel and JP-8 consist of thousands of chemical species and Arrhenius reaction steps. Oxygenated fuels such as bio-fuels and heavier hydrocarbons, such as from newer fossil fuel sources, are expected to have a much more complex chemistry requiring increasingly larger chemical kinetic models. Such models are beyond current computational capability, except for homogeneous or partially stirred reactor type calculations. The advent of highly parallel multi-core processors and graphical processing units (GPUs) promises a steep increase in computational performance in the coming years. This paper will present a software framework that translates the detailed chemical kinetic models to high-performance code targeted for GPU accelerators.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗