Search NASA⌕ Search

SEARCH · Search NASA

Results for “Communication costs”

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
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Scalable line and plane relaxation in a parallel structured multigrid solver

The efficient solution of sparse, linear systems that arise through the discretization of partial differential equations remains a key challenge for a range of high performance scientific simulations. One approach for reducing data movement and improving performance is by exposing and exploiting structure in a problem through the use of robust structured multilevel solvers. By choosing coarsening that preserves the structure of the problem, these methods maintain efficient structured computation and communication throughout the multigrid hierarchy. However, when coarsening is not permitted to be dependent on the operator, anisotropy must be addressed by the smoother — producing error compatible for coarse-grid correction with structured coarsening. Here, the components required in a scalable parallel structured solver are described with a focus on memory and communication efficiency of robust smoothers. While the implementation of communication and memory reduction techniques in smoothers integrated in a complete 3D solver present a significant engineering challenge, a novel approach is proposed that addresses these challenges systematically through a change to the solver’s execution model. Enabled by user-level threading paired with a set of data and communication abstractions, this approach permits seamless aggregation of communication in plane smoothers — directly reusing code for a 2D distributed multilevel cycle. Results show an effective reduction in communication costs for coarse-grid problems, and result in a speedup of 8.7x in smoothing routines shown in Fig. 12 using this approach. This produces a significant improvement to strong scalability while maintaining favorable weak scaling behavior. Finally, a parallel scaling study using a series of refined meshes is included that demonstrates the effectiveness of this approach in an application of interest.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Multistage mixed precision iterative refinement

Abstract Low precision arithmetic, in particular half precision (16‐bit) floating point arithmetic, is now available in commercial hardware. Using lower precision can offer significant savings in computation and communication costs with proportional savings in energy. Motivated by this, there has been a renewed interest in mixed precision iterative refinement schemes for solving linear systems , and new variants of GMRES‐based iterative refinement have been developed. Each particular variant with a given combination of precisions leads to different condition number‐based constraints for convergence of the backward and forward errors, and each has different performance costs. The constraints for convergence given in the literature are, as an artifact of the analyses, often overly strict in practice, and thus could lead a user to select a more expensive variant when a less expensive one would have sufficed. In this work, we develop a multistage mixed precision iterative refinement solver which aims to combine existing mixed precision approaches to balance performance and accuracy and improve usability. For a user‐specified initial combination of precisions, the algorithm begins with the least expensive approach and convergence is monitored via inexpensive computations with quantities produced during the iteration. If slow convergence or divergence is detected using particular stopping criteria, the algorithm switches to use a more expensive, but more reliable variant. A novel aspect of our approach is that, unlike existing implementations, our algorithm first attempts to use “stronger” GMRES‐based solvers for the solution update before resorting to increasing the precision(s). In some scenarios, this can avoid the need to refactorize the matrix in higher precision. We perform extensive numerical experiments on a variety of random dense problems and problems from real applications which confirm the benefits of the multistage approach.

Oktay, Eda↗
DOE OSTI2021

Parallel physics-informed neural networks via domain decomposition

Here we develop a distributed framework for the physics-informed neural networks (PINNs) based on two recent extensions, namely conservative PINNs (cPINNs) and extended PINNs (XPINNs), which employ domain decomposition in space and in time-space, respectively. This domain decomposition endows cPINNs and XPINNs with several advantages over the vanilla PINNs, such as parallelization capacity, large representation capacity, efficient hyperparameter tuning, and is particularly effective for multi-scale and multi-physics problems. Here, we present a parallel algorithm for cPINNs and XPINNs constructed with a hybrid programming model described by MPI + X, where X ∈ {CPUs, GPUs}. The main advantage of cPINN and XPINN over the more classical data and model parallel approaches is the flexibility of optimizing all hyperparameters of each neural network separately in each subdomain. We compare the performance of distributed cPINNs and XPINNs for various forward problems, using both weak and strong scalings. Our results indicate that for space domain decomposition, cPINNs are more efficient in terms of communication cost but XPINNs provide greater flexibility as they can also handle time-domain decomposition for any differential equations, and can deal with any arbitrarily shaped complex subdomains. To this end, we also present an application of the parallel XPINN method for solving an inverse diffusion problem with variable conductivity on the United States map, using ten regions as subdomains.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2021

Toward exascale whole-device modeling of fusion devices: Porting the GENE gyrokinetic microturbulence code to GPU

GENE solves the five-dimensional gyrokinetic equations to simulate the development and evolution of plasma microturbulence in magnetic fusion devices. The plasma model used is close to first principles and computationally very expensive to solve in the relevant physical regimes. In order to use the emerging computational capabilities to gain new physics insights, several new numerical and computational developments are required. Here, we focus on the fact that it is crucial to efficiently utilize GPUs (graphics processing units) that provide the vast majority of the computational power on such systems. In this paper, we describe the various porting approaches considered and given the constraints of the GENE code and its development model, justify the decisions made, and describe the path taken in porting GENE to GPUs. We introduce a novel library called gtensor that was developed along the way to support the process. Performance results are presented for the ported code, which in a single node of the Summit supercomputer achieves a speed-up of almost 15× compared to running on central processing unit (CPU) only. Typical GPU kernels are memory-bound, achieving about 90% of peak. Our analysis shows that there is still room for improvement if we can refactor/fuse kernels to achieve higher arithmetic intensity. We also performed a weak parallel scalability study, which shows that the code runs well on a massively parallel system, but communication costs start becoming a significant bottleneck.

Germaschewski, K. (ORCID:0000000284956354)↗
DOE OSTI2020

Machine-learning prediction for quasiparton distribution function matrix elements

There have been rapid developments in the direct calculation in lattice QCD (LQCD) of the Bjorken-x dependence of hadron structure through large-momentum effective theory (LaMET). LaMET overcomes the previous limitation of LQCD to moments (that is, integrals over Bjorken x) of hadron structure, allowing LQCD to directly provide the kinematic regions where the experimental values are least known. LaMET requires large-momentum hadron states to minimize its systematics and allow us to reach small- x reliably. This means that very fine lattice spacing to minimize lattice artifacts at order $(P_za)^n$ will become crucial for next-generation LaMET-like structure calculations. Furthermore, such calculations require operators with long Wilson-link displacements, especially in finer lattice units, increasing the communication costs relative to that of the propagator inversion. In this work, we explore whether machine-learning algorithms can make predictions of correlators to reduce the computational cost of these LQCD calculations. We consider two algorithms, gradient-boosting decision tree and linear models, applied to LaMET data, the matrix elements needed to determine the kaon and $η_s$ unpolarized parton distribution functions (PDFs), meson distribution amplitude (DA), and the nucleon gluon PDF. We find that both algorithms can reliably predict the target observables with different prediction accuracy and systematic errors. The predictions from smaller displacement z to larger ones work better than those for momentum p due to the higher correlation among the data.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

Interface learning of multiphysics and multiscale systems

Complex natural or engineered systems comprise multiple characteristic scales, multiple spatiotemporal domains, and even multiple physical closure laws. To address such challenges, we introduce an interface learning paradigm and put forth a data-driven closure approach based on memory embedding to provide physically correct boundary conditions at the interface. To enable the interface learning for hyperbolic systems by considering the domain of influence and wave structures into account, we put forth the concept of upwind learning toward a physics-informed domain decomposition. The promise of the proposed approach is shown for a set of canonical illustrative problems. Here, we highlight that high-performance computing environments can benefit from this methodology to reduce communication costs among processing units in emerging machine-learning-ready heterogeneous platforms toward exascale era.

42 ENGINEERING↗
DOE OSTI2020

Optimizing High Performance Markov Clustering for Pre-Exascale Architectures

HipMCL is a high-performance distributed memory implementation of the popular Markov Cluster Algorithm (MCL) and can cluster large-scale networks within hours using a few thousand CPU-equipped nodes. It relies on sparse matrix computations and heavily makes use of the sparse matrix-sparse matrix multiplication kernel (SpGEMM). The existing parallel algorithms in HipMCL are not scalable to Exascale architectures, both due to their communication costs dominating the runtime at large concurrencies and also due to their inability to take advantage of accelerators that are increasingly popular. In this work, we systematically remove scalability and performance bottlenecks of HipMCL. We enable GPUs by performing the expensive expansion phase of the MCL algorithm on GPU. Additionally, we propose a CPU-GPU joint distributed SpGEMM algorithm called pipelined Sparse SUMMA and integrate a probabilistic memory requirement estimator that is fast and accurate. Furthermore, we develop a new merging algorithm for the incremental processing of partial results produced by the GPUs, which improves the overlap efficiency and the peak memory usage. We also integrate a recent and faster algorithm for performing SpGEMM on CPUs. We validate our new algorithms and optimizations with extensive evaluations. With the enabling of the GPUs and integration of new algorithms, HipMCL is up to 12.4x faster, being able to cluster a network with 70 million proteins and 68 billion connections just under 15 minutes using 1024 nodes of ORNL's Summit supercomputer.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Distributed Hierarchical Contour Trees

Contour trees are a significant tool for data analysis as they capture both local and global variation. However, their utility has been limited by scalability, in particular for distributed computation and storage. We report a distributed data structure for storing the contour tree of a data set distributed on a cluster, based on a fan-in hierarchy, and an algorithm for computing it based on the boundary tree that represents only the superarcs of a contour tree that involve contours that cross boundaries between blocks. This allows us to limit the communication cost for contour tree computation to the complexity of the block boundaries rather than of the entire data set.

Carr, Hamish A↗
DOE OSTI2024

Parallel Randomized Tucker Decomposition Algorithms

The Tucker tensor decomposition is a natural extension of the singular value decomposition (SVD) to multiway data. Here, we propose to accelerate Tucker tensor decomposition algorithms by using randomization and parallelization. We present two algorithms that scale to large data and many processors, significantly reduce both computation and communication cost compared to previous deterministic and randomized approaches, and obtain nearly the same approximation errors. The key idea in our algorithms is to perform randomized sketches with Kronecker-structured random matrices, which reduces computation compared to unstructured matrices and can be implemented using a fundamental tensor computational kernel. We provide probabilistic error analysis of our algorithms and implement a new parallel algorithm for the structured randomized sketch. Our experimental results demonstrate that our combination of randomization and parallelization achieves accurate Tucker decompositions much faster than alternative approaches. We observe up to a 16X speedup over the fastest deterministic parallel implementation on 3D simulation data.

Tucker decompositions↗
DOE OSTI2024

Two-Stage Estimation and Variance Modeling for Latency-Constrained Variational Quantum Algorithms

The quantum approximate optimization algorithm (QAOA) has enjoyed increasing attention in noisy, intermediate-scale quantum computing with its application to combinatorial optimization problems. QAOA has the potential to demonstrate a quantum advantage for NP-hard combinatorial optimization problems. As a hybrid quantum-classical algorithm, the classical component of QAOA resembles a simulation optimization problem in which the simulation outcomes are attainable only through a quantum computer. The simulation that derives from QAOA exhibits two unique features that can have a substantial impact on the optimization process: (i) the variance of the stochastic objective values typically decreases in proportion to the optimality gap, and (ii) querying samples from a quantum computer introduces an additional latency overhead. In this paper, we introduce a novel stochastic trust-region method derived from a derivative-free, adaptive sampling trust-region optimization method intended to efficiently solve the classical optimization problem in QAOA by explicitly taking into account the two mentioned characteristics. The key idea behind the proposed algorithm involves constructing two separate local models in each iteration: a model of the objective function and a model of the variance of the objective function. Exploiting the variance model allows us to restrict the number of communications with the quantum computer and also helps navigate the nonconvex objective landscapes typical in QAOA optimization problems. In conclusion, we numerically demonstrate the superiority of our proposed algorithm using the SimOpt library and Qiskit when we consider a metric of computational burden that explicitly accounts for communication costs.

Derivative-free Optimization↗
DOE OSTI2020

Measuring Atmospheric Gas Using Small Satellites (Final Report)

The research focused on demonstrating that a small satellite (< 10 kg) is able to provide useable data in support of stratospheric circulation and broader atmospheric measurement generally thought to required large, lengthy and high cost space programs. The research goals targeting enabling CubeSats to serve as science platforms including low-cost communication infrastructure, science enabling solar panels and a ruggedized small form factor science instrument were highly successful with demonstrated on-orbit results. The collection of science data on-orbit was not successful due to a high tumbling rate impacted to the spacecraft at time of release on orbit, preventing reliable solar charging and leading to premature ending of the mission. The successes acquired on orbit will act as a springboard for further research and space projects at LLNL. Several research groups (both private and government) have expressed interest in the area of this research that proved successful such as the solar panels and iridium crosslink as well as the LLNL developed CubeSat standard.

42 ENGINEERING↗
DOE OSTI2020

Variable Precision Computing (VPC) (Final Report)

The overall goal of the Variable Precision Computing project is exploring and identifying ways to reduce internodal communication cost in production physics simulation experiments running on a distributed-memory systems by compressing data, either in a lossy or lossless way, thus resulting in a significant computational performance increase. Among the several approaches proposed, the one discussed here has been proposed, designed and implemented by me. It concerns the application of information theory-derived metrics to classical molecular dynamics and computational fluid dynamics simulation experiments.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1978

Automated meteorological data from commercial aircraft via satellite: Present experience and future implications

A low-cost communications system to provide meteorological data from commercial aircraft, in neat real-time, on a fully automated basis has been developed. The complete system including the low profile antenna and all installation hardware weighs 34 kg. The prototype system was installed on a B-747 aircraft and provided meteorological data (wind angle and velocity, temperature, altitude and position as a function of time) on a fully automated basis. The results were exceptional. This concept is expected to have important implications for operational meteorology and airline route forecasting.

Steinberg, R.↗
NASA NTRS1978

Automated meteorological data from commercial aircraft via satellite - Present experience and future implications

The National Aeronautics and Space Administration has developed a low-cost communications system to provide meteorological data from commercial aircraft, in near real-time, on a fully automated basis. The complete system including the low profile antenna and all installation hardware weighs 34 kg. The prototype system has been installed on a Pan American B-747 aircraft and has been providing meteorological data (wind angle and velocity, temperature, altitude and position as a function of time) on a fully automated basis for the past several months. The results have been exceptional. This concept is expected to have important implications for operational meteorology and airline route forecasting.

Steinberg, R.↗
NASA NTRS1986

Simultaneous iterations algorithm for general eigenvalue problems on parallel processors

The method of simultaneous iteration with shift is extended to extraction of m-eigenpairs of a general eigenvalue problem of large order n in a parallel processing environment. The algorithm combines the power method and the Jacobi technique, and reduces to performing four basic operations. Parallel implementation of the algorithm is discussed in detail. The analysis accounts for computation and communication costs, and utilizes a parallel processing architecture of the ensemble type. Expressions for the computational efficiency and speedup are defined as a function of the problem and hardware parameters. Selected representative problems exhibit efficiencies ranging from 60 to 98 percent.

Utku, S.↗
NASA NTRS1988

Parallel matrix multiplication on the Connection Machine

Matrix multiplication is a computation and communication intensive problem. Six parallel algorithms for matrix multiplication on the Connection Machine are presented and compared with respect to their performance and processor usage. For n by n matrices, the algorithms have theoretical running times of O(n to the 2nd power log n), O(n log n), O(n), and O(log n), and require n, n to the 2nd power, n to the 2nd power, and n to the 3rd power processors, respectively. With careful attention to communication patterns, the theoretically predicted runtimes can indeed be achieved in practice. The parallel algorithms illustrate the tradeoffs between performance, communication cost, and processor usage.

Tichy, Walter F.↗
NASA NTRS1991

An efficient massively parallel Euler solver for unstructured grids

A data parallel mesh-vertex upwind finite-volume scheme for solving the Euler equations on triangular unstructured meshes is described. A novel vertex-based partitioning of the problem is introduced which minimizes the computation and communication costs associated with distributing the computation to the processors of a massively parallel computer. Finally, the performance of this unstructured computation on 8K processors of the Connection Machine CM-2 is compared with one processor of a Cray-YMP. The experiments show that 8K processors of the CM-2 achieve approximately 70 percent of the performance of one processor of the Cray-YMP on the unstructured mesh computations described here.

Hammond, Steven W.↗
NASA NTRS1990

On the suitability of the connection machine for direct particle simulation

The algorithmic structure was examined of the vectorizable Stanford particle simulation (SPS) method and the structure is reformulated in data parallel form. Some of the SPS algorithms can be directly translated to data parallel, but several of the vectorizable algorithms have no direct data parallel equivalent. This requires the development of new, strictly data parallel algorithms. In particular, a new sorting algorithm is developed to identify collision candidates in the simulation and a master/slave algorithm is developed to minimize communication cost in large table look up. Validation of the method is undertaken through test calculations for thermal relaxation of a gas, shock wave profiles, and shock reflection from a stationary wall. A qualitative measure is provided of the performance of the Connection Machine for direct particle simulation. The massively parallel architecture of the Connection Machine is found quite suitable for this type of calculation. However, there are difficulties in taking full advantage of this architecture because of lack of a broad based tradition of data parallel programming. An important outcome of this work has been new data parallel algorithms specifically of use for direct particle simulation but which also expand the data parallel diction.

Dagum, Leonard↗