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 73 records · Page 4

Real-time High-resolution X-Ray Computed Tomography

Computed Tomography (CT) serves as a key imaging technology that relies on computationally intensive filtering and back-projection algorithms for 3D image reconstruction. While conventional high-resolution image reconstruction (> 2K3) solutions provide quick results, they typically treat reconstruction as an offline workload to be performed remotely on large-scale HPC systems. The growing demand for post-construction AI-driven analytics and the need for real-time adjustments call for high-resolution reconstruction solutions that are feasible on local computing resources, i.e. a multi-GPU server at most. In this paper, we propose a novel approach that utilizes Tensor Cores to optimize image reconstruction without sacrificing precision. We also introduce a framework designed to enable real-time execution of end-to-end distributed image reconstruction in a multi-GPU environment. Evaluations conducted on a single Nvidia A100 and H100 GPU show performance improvements of 1.91 × and 2.15 × compared to highly optimized production libraries. Furthermore, our framework, when deployed on 8-card Nvidia A100 GPU system, demonstrates the ability to reconstruct real-world datasets into 20483 volumes (32 GB) in slightly more than one minute and 40963 volumes (256 GB) in 7 minutes.

Wu, Du↗

Porting Fragmentation Methods to Graphical Processing Units Using an OpenMP Application Programming Interface: Offloading the Fock Build for Low Angular Momentum Functions

Here, a framework to offload four-index two-electron repulsion integrals to graphical processing units (GPUs) using OpenMP is discussed. The method has been applied to the Fock build for low angular momentum s and p functions in both the restricted Hartree–Fock (RHF) and in the effective fragment molecular orbital (EFMO) framework. Benchmark calculations for the GPU code for the pure RHF method show an increasing speedup relative to the existing OpenMP CPU code in GAMESS from 1.04 to 52× for clusters of 70–569 water molecules. The parallel efficiency on 24 NVIDIA V100 GPU boards also increases when increasing the system size: from 75 to 94% for water clusters that contain 303–1120 molecules. In the EFMO framework, the GPU Fock build shows a high linear scalability up to 4608 V100s with a parallel efficiency of 96% for calculations on a solvated mesoporous silica nanoparticle system with ~67,000 basis functions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Massively Parallel Latent Semantic Analyzes using a Graphics Processing Unit

Latent Semantic Indexing (LSA) aims to reduce the dimensions of large Term-Document datasets using Singular Value Decomposition. However, with the ever expanding size of data sets, current implementations are not fast enough to quickly and easily compute the results on a standard PC. The Graphics Processing Unit (GPU) can solve some highly parallel problems much faster than the traditional sequential processor (CPU). Thus, a deployable system using a GPU to speedup large-scale LSA processes would be a much more effective choice (in terms of cost/performance ratio) than using a computer cluster. Due to the GPU s application-specific architecture, harnessing the GPU s computational prowess for LSA is a great challenge. We present a parallel LSA implementation on the GPU, using NVIDIA Compute Unified Device Architecture and Compute Unified Basic Linear Algebra Subprograms. The performance of this implementation is compared to traditional LSA implementation on CPU using an optimized Basic Linear Algebra Subprograms library. After implementation, we discovered that the GPU version of the algorithm was twice as fast for large matrices (1000x1000 and above) that had dimensions not divisible by 16. For large matrices that did have dimensions divisible by 16, the GPU algorithm ran five to six times faster than the CPU version. The large variation is due to architectural benefits the GPU has for matrices divisible by 16. It should be noted that the overall speeds for the CPU version did not vary from relative normal when the matrix dimensions were divisible by 16. Further research is needed in order to produce a fully implementable version of LSA. With that in mind, the research we presented shows that the GPU is a viable option for increasing the speed of LSA, in terms of cost/performance ratio.

99 GENERAL AND MISCELLANEOUS↗

ddcMD: A fully GPU-accelerated molecular dynamics program for the Martini force field

We have implemented the Martini force field within Lawrence Livermore National Laboratory’s molecular dynamics program, ddcMD. The program is extended to a heterogeneous programming model so that it can exploit graphics processing unit (GPU) accelerators. In addition to the Martini force field being ported to the GPU, the entire integration step, including thermostat, barostat, and constraint solver, is ported as well, which speeds up the simulations to 278-fold using one GPU vs one central processing unit (CPU) core. A benchmark study is performed with several test cases, comparing ddcMD and GROMACS Martini simulations. The average performance of ddcMD for a protein–lipid simulation system of 136k particles achieves 1.04 µs/day on one NVIDIA V100 GPU and aggregates 6.19 µs/day on one Summit node with six GPUs. The GPU implementation in ddcMD offloads all computations to the GPU and only requires one CPU core per simulation to manage the inputs and outputs, freeing up remaining CPU resources on the compute node for alternative tasks often required in complex simulation campaigns.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

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

In, a practical implementation of a novel adaptive precision block-Jacobi preconditioner is introduced. In particular, the authors present a heavily-tuned GPU implementation of the adaptive precision block-Jacobi preconditioner within the Ginkgo numerical linear algebra library. The performance of the methodology and implementation is demonstrated using the proposed preconditioning scheme within Ginkgo’s high-performance Conjugate Gradient (CG) implementation on an NVIDIA Volta GPU. In this report, we replicate a subset of the computational results presented in. The focus is generating results from Fig. 9 to evaluate the performance of using Ginkgo’s CG solver integrated with either the full or the adaptive precision block-Jacobi preconditioner applied to a variety of test cases

97 MATHEMATICS AND COMPUTING↗

MASSIVELY PARALLEL LATENT SEMANTIC ANALYSES USING A GRAPHICS PROCESSING UNIT

Latent Semantic Analysis (LSA) aims to reduce the dimensions of large term-document datasets using Singular Value Decomposition. However, with the ever-expanding size of datasets, current implementations are not fast enough to quickly and easily compute the results on a standard PC. A graphics processing unit (GPU) can solve some highly parallel problems much faster than a traditional sequential processor or central processing unit (CPU). Thus, a deployable system using a GPU to speed up large-scale LSA processes would be a much more effective choice (in terms of cost/performance ratio) than using a PC cluster. Due to the GPU’s application-specifi c architecture, harnessing the GPU’s computational prowess for LSA is a great challenge. We presented a parallel LSA implementation on the GPU, using NVIDIA® Compute Unifi ed Device Architecture and Compute Unifi ed Basic Linear Algebra Subprograms software. The performance of this implementation is compared to traditional LSA implementation on a CPU using an optimized Basic Linear Algebra Subprograms library. After implementation, we discovered that the GPU version of the algorithm was twice as fast for large matrices (1 000x1 000 and above) that had dimensions not divisible by 16. For large matrices that did have dimensions divisible by 16, the GPU algorithm ran fi ve to six times faster than the CPU version. The large variation is due to architectural benefi ts of the GPU for matrices divisible by 16. It should be noted that the overall speeds for the CPU version did not vary from relative normal when the matrix dimensions were divisible by 16. Further research is needed in order to produce a fully implementable version of LSA. With that in mind, the research we presented shows that the GPU is a viable option for increasing the speed of LSA, in terms of cost/performance ratio.

Cavanagh, J↗

Parallel Latent Semantic Analysis using a Graphics Processing Unit

Latent Semantic Analysis (LSA) can be used to reduce the dimensions of large Term-Document datasets using Singular Value Decomposition. However, with the ever expanding size of data sets, current implementations are not fast enough to quickly and easily compute the results on a standard PC. The Graphics Processing Unit (GPU) can solve some highly parallel problems much faster than the traditional sequential processor (CPU). Thus, a deployable system using a GPU to speedup large-scale LSA processes would be a much more effective choice (in terms of cost/performance ratio) than using a computer cluster. In this paper, we presented a parallel LSA implementation on the GPU, using NVIDIA Compute Unified Device Architecture (CUDA) and Compute Unified Basic Linear Algebra Subprograms (CUBLAS). The performance of this implementation is compared to traditional LSA implementation on CPU using an optimized Basic Linear Algebra Subprograms library. For large matrices that have dimensions divisible by 16, the GPU algorithm ran five to six times faster than the CPU version.

59 BASIC BIOLOGICAL SCIENCES↗

Batched sparse direct solver design and evaluation in SuperLU_DIST

Over the course of interactions with various application teams, the need for batched sparse linear algebra functions has emerged in order to make more efficient use of the GPUs for many small and sparse linear algebra problems. In this paper, we present our recent work on a batched sparse direct solver for GPUs. The sparse LU factorization is computed by the levels of the elimination tree, leveraging the batched dense operations at each level and a new batched Scatter GPU kernel. The sparse triangular solve is computed by the level sets of the directed acyclic graph (DAG) of the triangular matrix. Batched operations overcome the large overhead associated with launching many small kernels. For medium sized matrix batches with not-so-small bandwidth, using an NVIDIA A100 GPU, our new batched sparse direct solver is orders of magnitude faster than a batched banded solver and uses less than one-tenth of the memory.

Boukaram, Wajih↗

SU-E-J-60: Efficient Monte Carlo Dose Calculation On CPU-GPU Heterogeneous Systems

Purpose: It is well-known that the performance of GPU-based Monte Carlo dose calculation implementations is bounded by memory bandwidth. One major cause of this bottleneck is the random memory writing patterns in dose deposition, which leads to several memory efficiency issues on GPU such as un-coalesced writing and atomic operations. We propose a new method to alleviate such issues on CPU-GPU heterogeneous systems, which achieves overall performance improvement for Monte Carlo dose calculation. Methods: Dose deposition is to accumulate dose into the voxels of a dose volume along the trajectories of radiation rays. Our idea is to partition this procedure into the following three steps, which are fine-tuned for CPU or GPU: (1) each GPU thread writes dose results with location information to a buffer on GPU memory, which achieves fully-coalesced and atomic-free memory transactions; (2) the dose results in the buffer are transferred to CPU memory; (3) the dose volume is constructed from the dose buffer on CPU. We organize the processing of all radiation rays into streams. Since the steps within a stream use different hardware resources (i.e., GPU, DMA, CPU), we can overlap the execution of these steps for different streams by pipelining. Results: We evaluated our method using a Monte Carlo Convolution Superposition (MCCS) program and tested our implementation for various clinical cases on a heterogeneous system containing an Intel i7 quad-core CPU and an NVIDIA TITAN GPU. Comparing with a straightforward MCCS implementation on the same system (using both CPU and GPU for radiation ray tracing), our method gained 2-5X speedup without losing dose calculation accuracy. Conclusion: The results show that our new method improves the effective memory bandwidth and overall performance for MCCS on the CPU-GPU systems. Our proposed method can also be applied to accelerate other Monte Carlo dose calculation approaches. This research was supported in part by NSF under Grants CCF-1217906, and also in part by a research contract from the Sandia National Laboratories.

60 APPLIED LIFE SCIENCES↗

Graphics Processing Unit Enhanced Parallel Document Flocking Clustering

Analyzing and clustering documents is a complex problem. One explored method of solving this problem borrows from nature, imitating the flocking behavior of birds. One limitation of this method of document clustering is its complexity O(n2). As the number of documents grows, it becomes increasingly difficult to generate results in a reasonable amount of time. In the last few years, the graphics processing unit (GPU) has received attention for its ability to solve highly-parallel and semi-parallel problems much faster than the traditional sequential processor. In this paper, we have conducted research to exploit this archi- tecture and apply its strengths to the flocking based document clustering problem. Using the CUDA platform from NVIDIA, we developed a doc- ument flocking implementation to be run on the NVIDIA GEFORCE GPU. Performance gains ranged from thirty-six to nearly sixty times improvement of the GPU over the CPU implementation.

Cui, Xiaohui↗

PeleC: An adaptive mesh refinement solver for compressible reacting flows

Reacting flow simulations for combustion applications require extensive computing capabilities. Leveraging the AMReX library, the Pele suite of combustion simulation tools targets the largest supercomputers available and future exascale machines. We introduce PeleC, the compressible solver in the Pele suite, and detail its capabilities, including complex geometry representation, chemistry integration, and discretization. We present a comparison of development efforts using both OpenACC and AMReX’s C++ performance portability framework for execution on multiple GPU architectures. We discuss relevant details that have allowed PeleC to achieve high performance and scalability. PeleC’s performance characteristics are measured through relevant simulations on multiple supercomputers. The success of PeleC’s design for exascale is exhibited through demonstration of a 160 billion cell simulation and weak scaling onto 100% of Summit, an NVIDIA-based GPU supercomputer at Oak Ridge National Laboratory. Our results provide confidence that PeleC will enable future combustion science simulations with unprecedented fidelity.

97 MATHEMATICS AND COMPUTING↗

Enabling Combustion Science Simulations for Future Exascale Machines

Reacting flow simulations for combustion applications require extensive computing capabilities. Leveraging the AMReX library, the Pele suite of combustion simulation tools targets the largest supercomputers available and future exascale machines. We introduce PeleC, the compressible solver in the Pele suite, and detail its capabilities, including complex geometry representation, chemistry integration, and discretization. We present a comparison of development efforts using both OpenACC and AMReX's C++ performance portability framework for execution on multiple GPU architectures. We discuss relevant details that have allowed PeleC to achieve high performance and scalability. PeleC's performance characteristics are measured through relevant simulations on multiple supercomputers. The success of PeleC's design for exascale is exhibited through demonstration of a 160 billion cell simulation and weak scaling onto 100\% of Summit, an NVIDIA-based GPU supercomputer at Oak Ridge National Laboratory. Our results provide confidence that PeleC will enable future combustion science simulations with unprecedented fidelity.

combustion↗

Kokkos GPU Compiler

The Kokkos Clang compiler is a version of the Clang C++ compiler that has been modified to perform targeted code generation for Kokkos constructs in the goal of generating highly optimized code and to provide semantic (domain) awareness throughout the compilation toolchain of these constructs such as parallel for and parallel reduce. This approach is taken to explore the possibilities of exposing the developer’s intentions to the underlying compiler infrastructure (e.g. optimization and analysis passes within the middle stages of the compiler) instead of relying solely on the restricted capabilities of C++ template metaprogramming. To date our current activities have focused on correct GPU code generation and thus we have not yet focused on improving overall performance. The compiler is implemented by recognizing specific (syntactic) Kokkos constructs in order to bypass normal template expansion mechanisms and instead use the semantic knowledge of Kokkos to directly generate code in the compiler’s intermediate representation (IR); which is then translated into an NVIDIA-centric GPU program and supporting runtime calls. In addition, by capturing and maintaining the higher-level semantics of Kokkos directly within the lower levels of the compiler has the potential for significantly improving the ability of the compiler to communicate with the developer in the terms of their original programming model/semantics.

Moss, Nicholas↗

Kokkos GPU Compiler

The Kokkos Clang compiler is a version of the Clang C++ compiler that has been modified to perform targeted code generation for Kokkos constructs in the goal of generating highly optimized code and to provide semantic (domain) awareness throughout the compilation toolchain of these constructs such as parallel for and parallel reduce. This approach is taken to explore the possibilities of exposing the developer’s intentions to the underlying compiler infrastructure (e.g. optimization and analysis passes within the middle stages of the compiler) instead of relying solely on the restricted capabilities of C++ template metaprogramming. To date our current activities have focused on correct GPU code generation and thus we have not yet focused on improving overall performance. The compiler is implemented by recognizing specific (syntactic) Kokkos constructs in order to bypass normal template expansion mechanisms and instead use the semantic knowledge of Kokkos to directly generate code in the compiler’s intermediate representation (IR); which is then translated into an NVIDIA-centric GPU program and supporting runtime calls. In addition, by capturing and maintaining the higher-level semantics of Kokkos directly within the lower levels of the compiler has the potential for significantly improving the ability of the compiler to communicate with the developer in the terms of their original programming model/semantics.

Moss, Nicholas↗

SU-D-206-01: Employing a Novel Consensus Optimization Strategy to Achieve Iterative Cone Beam CT Reconstruction On a Multi-GPU Platform

Purpose: While compressed sensing-based cone-beam CT (CBCT) iterative reconstruction techniques have demonstrated tremendous capability of reconstructing high-quality images from undersampled noisy data, its long computation time still hinders wide application in routine clinic. The purpose of this study is to develop a reconstruction framework that employs modern consensus optimization techniques to achieve CBCT reconstruction on a multi-GPU platform for improved computational efficiency. Methods: Total projection data were evenly distributed to multiple GPUs. Each GPU performed reconstruction using its own projection data with a conventional total variation regularization approach to ensure image quality. In addition, the solutions from GPUs were subject to a consistency constraint that they should be identical. We solved the optimization problem with all the constraints considered rigorously using an alternating direction method of multipliers (ADMM) algorithm. The reconstruction framework was implemented using OpenCL on a platform with two Nvidia GTX590 GPU cards, each with two GPUs. We studied the performance of our method and demonstrated its advantages through a simulation case with a NCAT phantom and an experimental case with a Catphan phantom. Result: Compared with the CBCT images reconstructed using conventional FDK method with full projection datasets, our proposed method achieved comparable image quality with about one third projection numbers. The computation time on the multi-GPU platform was ∼55 s and ∼ 35 s in the two cases respectively, achieving a speedup factor of ∼ 3.0 compared with single GPU reconstruction. Conclusion: We have developed a consensus ADMM-based CBCT reconstruction method which enabled performing reconstruction on a multi-GPU platform. The achieved efficiency made this method clinically attractive.

60 APPLIED LIFE SCIENCES↗

Deep Packet/Flow Analysis using GPUs

aDeep packet inspection (DPI) faces severe performance challenges in high-speed networks (40/100 GE) as it requires a large amount of raw computing power and high I/O throughputs. Recently, researchers have tentatively used GPUs to address the above issues and boost the performance of DPI. Typically, DPI applications involve highly complex operations in both per-packet and per-flow data level, often in real-time. The parallel architecture of GPUs fits exceptionally well for per-packet network traffic processing. However, for stateful network protocols such as TCP, their data stream need to be reconstructed in a per-flow level to deliver a consistent content analysis. Since the flow-centric operations are naturally anti-parallel and often require large memory space for buffering out-of-sequence packets, they can be problematic for GPUs, whose memory is normally limited to several gigabytes. In this work, we present a highly efficient GPU-based deep packet/flow analysi s framework. The proposed design includes a purely GPU-implemented flow tracking and TCP stream reassembly. Instead of buffering and waiting for TCP packets to become in sequence, our framework process the packets in batch and uses a deterministic finite automaton (DFA) with prefix-/suffix- tree method to detect patterns across out-of-sequence packets that happen to be located in different batches. Evaluation shows that our code can reassemble and forward tens of millions of packets per second and conduct a stateful signature-based deep packet inspection at 55 Gbit/s using an NVIDIA K40 GPU.

Gong, Qian↗

Deep Packet/Flow Analysis using GPUs

Deep packet inspection (DPI) faces severe performance challenges in high-speed networks (40/100 GE) as it requires a large amount of raw computing power and high I/O throughputs. Recently, researchers have tentatively used GPUs to address the above issues and boost the performance of DPI. Typically, DPI applications involve highly complex operations in both per-packet and per-flow data level, often in real-time. The parallel architecture of GPUs fits exceptionally well for per-packet network traffic processing. However, for stateful network protocols such as TCP, their data stream need to be reconstructed in a per-flow level to deliver a consistent content analysis. Since the flow-centric operations are naturally antiparallel and often require large memory space for buffering out-of-sequence packets, they can be problematic for GPUs, whose memory is normally limited to several gigabytes. In this work, we present a highly efficient GPU-based deep packet/flow analysis framework. The proposed design includes a purely GPU-implemented flow tracking and TCP stream reassembly. Instead of buffering and waiting for TCP packets to become in sequence, our framework process the packets in batch and uses a deterministic finite automaton (DFA) with prefix-/suffix- tree method to detect patterns across out-of-sequence packets that happen to be located in different batches. In conclusion, evaluation shows that our code can reassemble and forward tens of millions of packets per second and conduct a stateful signature-based deep packet inspection at 55 Gbit/s using an NVIDIA K40 GPU.

97 MATHEMATICS AND COMPUTING↗