Search NASA⌕ Search

SEARCH · Search NASA

Results for “NVIDIA GPU”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2017

GPU-Centric Communication on NVIDIA GPU Clusters with InfiniBand: A Case Study with OpenSHMEM

GPUs have become an essential component for building compute clusters with high compute density and high performance per watt. As such clusters scale to have 1000s of GPUs, efficiently moving data between the GPUs becomes imperative to get maximum performance. NVSHMEM is an implementation of the OpenSHMEM standard for NVIDIA GPU clusters which allows communication to be issued from inside GPU kernels. In earlier work, we have shown how NVSHMEM can be used to achieve better application performance on GPUs connected through PCIe or NVLink. As part of this effort, we implement IB verbs for Mellanox InfiniBand adapters in CUDA. We evaluate different design alternatives, taking into consideration the relaxed memory model, automatic memory access coalescing and thread hierarchy on the GPU. We also consider correctness issues that arise in these designs. We take advantage of these designs transparently or through API extensions in NVSHMEM. With micro-benchmarks, we show that a Nvidia Pascal P100 GPU is able saturate the network bandwidth using only one or two of its 56 available streaming multiprocessors (SM). On a single GPU using a single IB EDR adapter, we achieve a throughput of around 90 million messages per second. In addition, we implement a 2dstencil application kernel using NVSHMEM and compare its performance with a CUDA-aware MPI-based implementation that uses GPUDirect RDMA. Speedups in the range of 23% to 42% are seen for input sizes large enough to fill the occupancy of Nvidia Pascal P100 GPUs on 2 to 4 nodes indicating that there are gains to be had by eliminating the CPU from the communication path when all computation runs on the GPU.

Potluri, Sreeram↗
DOE OSTI2022

Understanding Performance Portability of Bioinformatics Applications in SYCL on an NVIDIA GPU

Our goal is to have a better understanding of performance portability of SYCL kernels on a GPU. Toward this goal, we migrate representative kernels in bioinformatics applications from CUDA to SYCL, evaluate their performance on an NVIDIA GPU, and explain the performance gaps through performance profiling and analyses. We hope that the findings provide valuable feedback to the development of the SYCL ecosystem.

Jin, Zheming↗
DOE OSTI2022

Performance portability study of epistasis detection using SYCL on NVIDIA GPU

We describe the experience of converting a CUDA implementation of a high-order epistasis detection algorithm to SYCL. The goals are for our work to be useful to application and compiler developers with a detailed description of migration paths between CUDA and SYCL. Evaluating the CUDA and SYCL applications on an NVIDIA V100 GPU, we find that the optimization of loop unrolling needs to be applied manually to the SYCL kernel for obtaining comparable performance. The performance of the SYCL group reduce function, an alternative to the CUDA warp-based reduction, depends on the problem and work group sizes. The 64-bit popcount operation implemented with tree of adders is slightly faster than the built-in popcount operation. When the number of OpenMP threads is four, the highest performance of the SYCL and CUDA applications are comparable.

Jin, Zheming↗
DOE OSTI2021

NWQ-sim

NWQSim is a quantum circuit simulation environment developed at PNNL. It currently includes two major components: a state-vector simulator (SV-Sim) and a density matrix simulator (DM-Sim) and we may add more components, such as a Clifford simulator, in the future effort. NWQSim has two language interface: C/C++ and Python. It supports Q#/QDK frontend through QIR and QIR-runtime. It supports Qiskit and Cirq frontends through OpenQASM. NWQSim runs on several backends: Intel-CPU, Intel-Xeon-Phi, AMD-CPU, AMD-GPU, NVIDIA-GPU, and IBM-CPU. It supports three modes: (1) single processor, such as a single CPU (with and without AVX2 and AVX512 acceleration), a single NVIDIA GPU or a single AMD GPU; (2) single-node-multi-processors, such as multi-CPUs/Xeon-Phis, multi-NVIDA/AMD GPUs; (3) multi-nodes, such as a CPU cluster, a Xeon-Phi cluster (e.g., ANL Theta, NERSC Cori), an NVIDIA cluster (e.g., ORNL Summit, NERSC Perlmutter).

Li, Ang↗
DOE OSTI2017

Modern gyrokinetic particle-in-cell simulation of fusion plasmas on top supercomputers

The Gyrokinetic Toroidal Code at Princeton (GTC-P) is a highly scalable and portable particle-in-cell (PIC) code. It solves the 5D Vlasov-Poisson equation featuring efficient utilization of modern parallel computer architectures at the petascale and beyond. Motivated by the goal of developing a modern code capable of dealing with the physics challenge of increasing problem size with sufficient resolution, new thread-level optimizations have been introduced as well as a key additional domain decomposition. GTC-P's multiple levels of parallelism, including inter-node 2D domain decomposition and particle decomposition, as well as intra-node shared memory partition and vectorization have enabled pushing the scalability of the PIC method to extreme computational scales. In this paper, we describe the methods developed to build a highly parallelized PIC code across a broad range of supercomputer designs. This particularly includes implementations on heterogeneous systems using NVIDIA GPU accelerators and Intel Xeon Phi (MIC) co-processors and performance comparisons with state-of-the-art homogeneous HPC systems such as Blue Gene/Q. New discovery science capabilities in the magnetic fusion energy application domain are enabled, including investigations of Ion-Temperature-Gradient (ITG) driven turbulence simulations with unprecedented spatial resolution and long temporal duration. Performance studies with realistic fusion experimental parameters are carried out on multiple supercomputing systems spanning a wide range of cache capacities, cache-sharing configurations, memory bandwidth, interconnects and network topologies. Furthermore, these performance comparisons using a realistic discovery-science-capable domain application code provide valuable insights on optimization techniques across one of the broadest sets of current high-end computing platforms worldwide.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2023

OpenACC Unified Programming Environment for Multi-hybrid Acceleration with GPU and FPGA

Accelerated computing in HPC such as with GPU, plays a central role in HPC nowadays. However, in some complicated applications with partially different performance behavior is hard to solve with a single type of accelerator where GPU is not the perfect solution in these cases. We are developing a framework and transpiler allowing the users to program the codes with a single notation of OpenACC to be compiled for multi-hybrid accelerators, named MHOAT (Multi-Hybrid OpenACC Translator) for HPC applications. MHOAT parses the original code with directives to identify the target accelerating devices, currently supporting NVIDIA GPU and Intel FPGA, dispatching these specific partial codes to background compilers such as NVIDIA HPC SDK for GPU and OpenARC research compiler for FPGA, then assembles binaries for the final object with FPGA bitstream file. In this paper, we present the concept, design, implementation, and performance evaluation of a practical astrophysics simulation code where we successfully enhanced the performance up to 10 times faster than the GPU-only solution.

Boku, Taisuke↗
DOE OSTI2022

Development and performance of a HemeLB GPU code for human-scale blood flow simulation

In recent years, it has become increasingly common for high performance computers (HPC) to possess some level of heterogeneous architecture - typically in the form of GPU accelerators. In some machines these are isolated within a dedicated partition, whilst in others they are integral to all compute nodes - often with multiple GPUs per node - and provide the majority of a machine's compute performance. In light of this trend, it is becoming essential that codes deployed on HPC are updated to execute on accelerator hardware. Here, in this paper, we introduce a GPU implementation of the 3D blood flow simulation code HemeLB that has been developed using CUDA C++. We demonstrate how taking advantage of NVIDIA GPU hardware can achieve significant performance improvements compared to the equivalent CPU only code on which it has been built whilst retaining the excellent strong scaling characteristics that have been repeatedly demonstrated by the CPU version. With HPC positioned on the brink of the exascale era, we use HemeLB as a motivation to provide a discussion on some of the challenges that many users will face when deploying their own applications on upcoming exascale machines.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Fast GPU-Based Generation of Large Graph Networks From Degree Distributions

Synthetically generated, large graph networks serve as useful proxies to real-world networks for many graph-based applications. The ability to generate such networks helps overcome several limitations of real-world networks regarding their number, availability, and access. Here, we present the design, implementation, and performance study of a novel network generator that can produce very large graph networks conforming to any desired degree distribution. The generator is designed and implemented for efficient execution on modern graphics processing units (GPUs). Given an array of desired vertex degrees and number of vertices for each desired degree, our algorithm generates the edges of a random graph that satisfies the input degree distribution. Multiple runtime variants are implemented and tested: 1) a uniform static work assignment using a fixed thread launch scheme, 2) a load-balanced static work assignment also with fixed thread launch but with cost-aware task-to-thread mapping, and 3) a dynamic scheme with multiple GPU kernels asynchronously launched from the CPU. The generation is tested on a range of popular networks such as Twitter and Facebook, representing different scales and skews in degree distributions. Results show that, using our algorithm on a single modern GPU (NVIDIA Volta V100), it is possible to generate large-scale graph networks at rates exceeding 50 billion edges per second for a 69 billion-edge network. GPU profiling confirms high utilization and low branching divergence of our implementation from small to large network sizes. For networks with scattered distributions, we provide a coarsening method that further increases the GPU-based generation speed by up to a factor of 4 on tested input networks with over 45 billion edges.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

BinFPE

BinFPE is a tool that detects floating-point exceptions (NaN, infinity, and subnormal quantities) in NVIDIA GPU applications using binary instrumentation. It requires no re-compilation of the application and can analyze libraries. The tool extends NVBit (https://github.com/NVlabs/NVBit), which is provided by NVIDIA Labs to analyze binaries. When the tool detects a floating-point exception it reports to the user the location of the exception (file and line number).

Laguna Peralta, Ignacio↗
DOE OSTI2020

Indicator-directed Dynamic Power Management for Iterative Workloads on GPU-Accelerated Systems

Modern high-performance and warehouse computing centers show strong interest in minimizing system power consumption while satisfying customers’ quality of service (QoS). Dynamic voltage and frequency scaling (DVFS) is effective for achieving this goal. Nevertheless, automating the process online and making it transparent to users must address three major challenges: (1) Complexity — today’s hardware components (e.g., CPUs, GPUs, memory, network, etc.) can be configured in several or dozens of frequency/voltage states for satisfying divergent system demands. Given their combination and the emergence of heterogeneity, searching the optimal configuration in the design space online can be timing consuming. (2) QoS guarantee — user-defined objectives such as power constraint and performance target must be monitored, predicted and ensured at the best effort. (3) Adaptability — various known and unknown workloads run on systems. Workloads characteristics should be quickly determined and configurations dynamically adjusted in accord with workloads and QoS. In this work, we focus on applications exhibiting an interesting feature – iterative or periodic, which is common among conventional HPC and emerging machine learning workloads. We propose an online dynamic power-performance (ODPP) management framework to dynamically adjust GPU DVFS configurations to meet performance and power objectives and constraints, without any code annotation or intrusion. Particularly, ODPP extracts the performance and power indicators for applications from their resources utilization profiles in a short episode. It further automatically constructs an accurate model that infers from the indicators how the application's performance and power vary with GPU core and memory frequencies. Aided with the model, for both seen and unseen applications, ODPP can quickly determine the most appropriate DVFS configuration for their execution. We evaluate ODPP on an NVIDIA GPU using multiple exascale computing (ECP) and deep learning applications.

Zou, Pengfei↗
DOE OSTI2021

Parallelizing the Unpacking and Clustering of Detector Data for Reconstruction of Charged Particle Tracks on Multi-core CPUs and Many-core GPUs

We present results from parallelizing the unpacking and clustering steps of the raw data from the silicon strip modules for reconstruction of charged particle tracks. Throughput is further improved by concurrently processing multiple events using nested OpenMP parallelism on CPU or CUDA streams on GPU. The new implementation along with earlier work in developing a parallelized and vectorized implementation of the combinatoric Kalman filter algorithm has enabled efficient global reconstruction of the entire event on modern computer architectures. We demonstrate the performance of the new implementation on Intel Xeon and NVIDIA GPU architectures.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2017

Locality-Aware CTA Clustering For Modern GPUs

In this paper, we proposed a novel clustering technique for tapping into the performance potential of a largely ignored type of locality: inter-CTA locality. We first demonstrated the capability of the existing GPU hardware to exploit such locality, both spatially and temporally, on L1 or L1/Tex unified cache. To verify the potential of this locality, we quantified its existence in a broad spectrum of applications and discussed its sources of origin. Based on these insights, we proposed the concept of CTA-Clustering and its associated software techniques. Finally, We evaluated these techniques on all modern generations of NVIDIA GPU architectures. The experimental results showed that our proposed clustering techniques could significantly improve on-chip cache performance.

Data Locality↗
DOE OSTI2021

Replicated Computational Results (RCR) Report for “Adaptive Precision Block-Jacobi for High Performance Preconditioning in the Ginkgo Linear Algebra Software”

The article by Flegar et al. titled “Adaptive Precision Block-Jacobi for High Performance Preconditioning in the Ginkgo Linear Algebra Software” presents a novel, practical implementation of an adaptive precision block-Jacobi preconditioner. Performance results using state-of-the-art GPU architectures for the block-Jacobi preconditioner generation and application demonstrate the practical usability of the method, compared to a traditional full-precision block-Jacobi preconditioner. A production-ready implementation is provided in the Ginkgo numerical linear algebra library. In this report, the Ginkgo library is reinstalled and performance results are generated to perform a comparison to the original results when using Ginkgo’s Conjugate Gradient solver with either the full or the adaptive precision block-Jacobi preconditioner for a suite of test problems on an NVIDIA GPU accelerator. After completing this process, the published results are deemed reproducible.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

A single-tree algorithm to compute the Euclidean minimum spanning tree on GPUs

Computing the Euclidean minimum spanning tree (Emst) is a computationally demanding step of many algorithms. While work-efficient serial and multithreaded algorithms for computing Emst are known, designing an efficient GPU algorithm is challenging due to a complex branching structure, data dependencies, and load imbalances. In this paper, we propose a single-tree $Bor\mathring{u}vka$-based algorithm for computing Emst on GPUs. We use an efficient nearest neighbor algorithm and reduce the number of the required distance calculations by avoiding traversing subtrees with leaf nodes in the same component. The developed algorithms are implemented in a performance portable way using ArborX, an open-source geometric search library based on the Kokkos framework. We evaluate the proposed algorithm on various 2D and 3D datasets, show and compare it with the current state-of-the-art open-source CPU implementations. We demonstrate 4-24 × speedup over the fastest multi-threaded implementation. We prove the portability of our implementation by providing results on a variety of hardware: AMD EPYC 7763, Nvidia A100 and AMD MI250X. We show scalability of the implementation, computing Emst for 37 million 3D cosmological dataset in under a 0.5 second on a single A100 Nvidia GPU.

Prokopenko, Andrey↗
DOE OSTI2021

Statistically-informed deep learning for gravitational wave parameter estimation

We introduce deep learning models to estimate the masses of the binary components of black hole mergers, $(m_1,m_2)$, and three astrophysical properties of the post-merger compact remnant, namely, the final spin, $a_\mathrm f$, and the frequency and damping time of the ringdown oscillations of the fundamental $\ell = m = 2$ bar mode, $(\omega_\mathrm R, \omega_\mathrm I)$. Our neural networks combine a modified WaveNet architecture with contrastive learning and normalizing flow. We validate these models against a Gaussian conjugate prior family whose posterior distribution is described by a closed analytical expression. Upon confirming that our models produce statistically consistent results, we used them to estimate the astrophysical parameters $(m_1,m_2, a_\mathrm f, \omega_\mathrm R, \omega_\mathrm I)$ of five binary black holes: GW150914, GW170104, GW170814, GW190521 and GW190630. We use PyCBC Inference to directly compare traditional Bayesian methodologies for parameter estimation with our deep learning based posterior distributions. Our results show that our neural network models predict posterior distributions that encode physical correlations, and that our data-driven median results and 90% confidence intervals are similar to those produced with gravitational wave Bayesian analyses. This methodology requires a single V100 NVIDIA GPU to produce median values and posterior distributions within two milliseconds for each event. Furthermore, this neural network, and a tutorial for its use, are available at the Data and Learning Hub for Science.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2017

Möbius domain-wall fermions on gradient-flowed dynamical HISQ ensembles

We report on salient features of a mixed lattice QCD action using valence Möbius domain-wall fermions solved on the dynamical 𝑁 𝑓 = 2 + 1 + 1 highly improved staggered quark sea-quark ensembles generated by the MILC Collaboration. The approximate chiral symmetry properties of the valence fermions are shown to be significantly improved by utilizing the gradient-flow scheme to first smear the highly improved staggered quark configurations. The greater numerical cost of the Möbius domain-wall inversions is mitigated by the highly efficient QUDA library optimized for NVIDIA GPU accelerated compute nodes. We have created an interface to this optimized QUDA solver in Chroma. We provide tuned parameters of the action and performance of QUDA using ensembles with the lattice spacings 𝑎 ≃ {0.15,0.12,0.09} fm and pion masses 𝑚 𝜋 ≃ {310,220,130} MeV. We have additionally generated two new ensembles with 𝑎 ∼ 0.12 fm and 𝑚 𝜋 ∼ {400,350} MeV. Furthermore, with a fixed flow time of 𝑡 𝑔⁡𝑓 =1 in lattice units, the residual chiral symmetry breaking of the valence fermions is kept below 10% of the light quark mass on all ensembles, 𝑚 res ≲ 0.1 × 𝑚 𝑙 , with moderate values of the fifth dimension 𝐿 5 and a domain-wall height 𝑀 5 ≤ 1.3. As a benchmark calculation, we perform a continuum, infinite volume, physical pion and kaon mass extrapolation of 𝐹 𝐾 ±⁡ /𝐹 𝜋 ± and demonstrate our results are independent of flow time and consistent with the FLAG determination of this quantity at the level of less than one standard deviation.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2019

ExaTENSOR

ExaTENSOR is a software math library for performing basic numerical tensor algebra operations on distributed heterogeneous HPC platforms. The library provides the following basic numerical tensor algebra primitives: tensor contraction, tensor product, tensor addition, tensor scaling by a scalar, etc. ExaTENSOR supports dense, block-sparse and hierarchical block-sparse tensors stored and processed across many heterogeneous HPC nodes (node equipped with NVIDIA GPU and Intel Xeon Phi have been considered so far). The architecture of ExaTENSOR is based on the concept of domain-specific virtual processor (DSVP), that is, an intermediate software layer capable of processing domain-specific instructions, in this case, numerical tensor algebra instructions. In this way, a separation of the domain-specific algorithm expression and hardware-agnostic algorithm execution is achieved, thus ensuring portability of the applications which use ExaTENSOR as a numerical backend.

Liakh, Dmytro↗