Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Parallel communication”

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 OSTI2021

Logically Parallel Communication for Fast MPI+Threads Applications

Supercomputing applications are increasingly adopting the MPI+threads programming model over the traditional "MPI everywhere" approach to better handle the disproportionate increase in the number of cores compared with other on-node resources. In practice, however, most applications observe a slower performance with MPI+threads primarily because of poor communication performance. Recent research efforts on MPI libraries address this bottleneck by mapping logically parallel communication, that is, operations that are not subject to MPI's ordering constraints to the underlying network parallelism. Domain scientists, however, typically do not expose such communication independence information because the existing MPI-3.1 standard's semantics can be limiting. Researchers had initially proposed user-visible endpoints to combat this issue, but such a solution requires intrusive changes to the standard (new APIs). The upcoming MPI-4.0 standard, on the other hand, allows applications to relax unneeded semantics and provides them with many opportunities to express logical communication parallelism. Here in this article, we show how MPI+threads applications can achieve high performance with logically parallel communication. Through application case studies, we compare the capabilities of the new MPI-4.0 standard with those of the existing one and user-visible endpoints (upper bound). Logical communication parallelism can boost the overall performance of an application by over 2x.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Optimizing Irregular Communication with Neighborhood Collectives and Locality-Aware Parallelism

Irregular communication often limits both the performance and scalability of parallel applications. Typically, applications individually implement irregular communication as point-to-point, and any optimizations are integrated directly into the application. As a result, these optimizations lack portability. It is difficult to optimize point-to-point messages within MPI, as the interface for single messages provides no information on the collection of all communication to be performed. However, the persistent neighbor collective API, released in the MPI 4 standard, provides an interface for portable optimizations of irregular communication within MPI libraries. This paper presents methods for implementing existing optimizations for irregular communication within neighborhood collectives, analyzes the impact of replacing point-to-point communication in existing codebases such as Hypre BoomerAMG with neighborhood collectives, and finally shows up to a 1.38x speedup on sparse matrix-vector multiplication communication within a BoomerAMG solve through the use of our optimized neighbor collectives. Here, the authors analyze three implementations of persistent neighborhood collectives for Alltoallv: an unoptimized wrapper of standard point-to-point communication, and two locality-aware aggregating methods. The second locality-aware implementation exposes an non-standard interface to perform additional optimization, and the authors present the additional 0.07x speedup from the extended interface. All optimizations are available in an open-source codebase, MPI Advance, which sits on top of MPI, allowing for optimizations to be added into existing codebases regardless of the system MPI install.

AMG↗
DOE OSTI2020

Optimizing the hypre solver for manycore and GPU architectures

The solution of large-scale combustion problems with codes such as Uintah on modern computer architectures requires the use of multithreading and GPUs to achieve performance. Uintah uses a low-Mach number approximation that requires iteratively solving a large system of linear equations. The Hypre iterative solver has solved such systems in a scalable way for Uintah, but the use of OpenMP with Hypre leads to at least slowdown due to OpenMP overheads. The proposed solution uses the MPI Endpoints within Hypre, where each team of threads acts as a different MPI rank. This approach minimizes OpenMP synchronization overhead and performs as fast or (up to 1.44) faster than Hypre's MPI-only version, and allows the rest of Uintah to be optimized using OpenMP. The profiling of the GPU version of Hypre shows the bottleneck to be the launch overhead of thousands of micro-kernels. The GPU performance was improved by fusing these micro-kernels and was further optimized by using Cuda-aware MPI, resulting in an overall speedup of 1.16—1.44 compared to the baseline GPU implementation. The above optimization strategies were published in the International Conference on Computational Science 2020 [1]. This work extends the previously published research by carrying out the second phase of communication-centered optimizations in Hypre to improve its scalability on large-scale supercomputers. Additionally, this includes an efficient non-blocking inter-thread communication scheme, communication-reducing patch assignment, and expression of logical communication parallelism to a new version of the MPICH library that utilizes the underlying network parallelism [2]. The above optimizations avoid communication bottlenecks previously observed during strong scaling and improve performance by up to 2 on 256 nodes of Intel Knight's Landing processor.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Parallel Memory-Independent Communication Bounds for SYRK

In this paper, we focus on the parallel communication cost of multiplying a matrix with its transpose, known as a symmetric rank-k update (SYRK). SYRK requires half the computation of general matrix multiplication because of the symmetry of the output matrix. Recent work (Beaumont et al., SPAA '22) has demonstrated that the sequential I/O complexity of SYRK is also a constant factor smaller than that of general matrix multiplication. Inspired by this progress, we establish memory-independent parallel communication lower bounds for SYRK with smaller constants than general matrix multiplication, and we show that these constants are tight by presenting communication-optimal algorithms. The crux of the lower bound proof relies on extending a key geometric inequality to symmetric computations and analytically solving a constrained nonlinear optimization problem. Here, the optimal algorithms use a triangular blocking scheme for parallel distribution of the symmetric output matrix and corresponding computation.

Communication costs↗
DOE OSTI2021

Optimization Plugin Library

The Optimization Plugin library ("op") is a lightweight general optimization solver interface. The primary purpose of op is to simplify the process of integrating different optimization solvers (serial or parallel) with scalable parallel physics engines. By design it has several features that help make this a reality. The core abstraction interface was developed to encompass a large class of optimization problems in an optimizer-agnostic way. This enables us to describe the optimization problem once and then use a variety of supported "op" optimizers with ideally no code-changes. The abstraction interface is made up of lightweight wrappers that make it easy to integrate with existing simulation codes. This makes integration less intrusive and should minimize changes to existing physics codes. The "op" interface includes an assortment of utility methods that help specify parallel communication patterns as well as methods to convert from optimization-specific interfaces to the more general "op" interface. Lastly a dynamic library linking interface is provided to allow for use of proprietary optimization engines without explicit reference in the source code, along with standard shared library interfaces for opensource engines.

Jekel, CharlesF↗
DOE OSTI2023

Massively parallel modeling and inversion of electrical resistivity tomography data using PFLOTRAN

Abstract. Electrical resistivity tomography (ERT) is a broadly accepted geophysical method for subsurface investigations. Interpretation of field ERT data usually requires the application of computationally intensive forward modeling and inversion algorithms. For large-scale ERT data, the efficiency of these algorithms depends on the robustness, accuracy, and scalability on high-performance computing resources. In this regard, we present a robust and highly scalable implementation of forward modeling and inversion algorithms for ERT data. The implementation is publicly available and developed within the framework of PFLOTRAN, an open-source, state-of-the-art massively parallel subsurface flow and transport simulation code. The forward modeling is based on a finite-volume discretization of the governing differential equations, and the inversion uses a Gauss–Newton optimization scheme. To evaluate the accuracy of the forward modeling, two examples are first presented by considering layered (1D) and 3D earth conductivity models. The computed numerical results show good agreement with the analytical solutions for the layered earth model and results from a well-established code for the 3D model. Inversion of ERT data, simulated for a 3D model, is then performed to demonstrate the inversion capability by recovering the conductivity of the model. To demonstrate the parallel performance of PFLOTRAN's ERT process model and inversion capabilities, large-scale scalability tests are performed by using up to 131 072 processes on a leadership class supercomputer. These tests are performed for the two most computationally intensive steps of the ERT inversion: forward modeling and Jacobian computation. For the forward modeling, we consider models with up to 122 ×106 degrees of freedom (DOFs) in the resulting system of linear equations and demonstrate that the code exhibits almost linear scalability on up to 10 000 DOFs per process. On the other hand, the code shows superlinear scalability for the Jacobian computation, mainly because all computations are fairly evenly distributed over each process with no parallel communication.

58 GEOSCIENCES↗
DOE OSTI2022

Accelerating Collective Communication in Data Parallel Training across Deep Learning Frameworks

This work develops new techniques within Horovod, a generic communication library supporting data parallel training across deep learning frameworks. In particular, we improve the Horovod control plane by implementing a new coordination scheme that takes advantage of the characteristics of the typical data parallel training paradigm, namely the repeated execution of collectives on the gradients of a fixed set of tensors. Using a caching strategy, we execute Horovod’s existing coordinator-worker logic only once during a typical training run, replacing it with a more efficient decentralized orchestration strategy using the cached data and a global intersection of a bitvector for the remaining training duration. Next, we introduce a feature for end users to explicitly group collective operations, enabling finer grained control over the communication buffer sizes. To evaluate our proposed strategies, we conduct experiments on a world-class supercomputer — Summit. We compare our proposals to Horovod’s original design and observe 2x performance improvement at a scale of 6000 GPUs; we also compare them against tf.distribute and torch.DDP and achieve 12% better and comparable performance, respectively, using up to 1536 GPUs; we compare our solution against BytePS in typical HPC settings and achieve about 20% better performance on a scale of 768 GPUs. Finally, we test our strategies on a scientific application (STEMDL) using up to 27,600 GPUs (the entire Summit) and show that we achieve a near-linear scaling of 0.93 with a sustained performance of 1.54 exaflops (with standard error +- 0.02) in FP16 precision.

Romero, Joshua↗
DOE OSTI2020

3D Coded SUMMA: Communication-Efficient and Robust Parallel Matrix Multiplication

In this paper, we propose a novel fault-tolerant parallel matrix multiplication algorithm called 3D Coded SUMMA that achieves higher failure-tolerance than replication-based schemes for the same amount of redundancy. This work bridges the gap between recent developments in coded computing and fault-tolerance in high-performance computing (HPC). The core idea of coded computing is the same as algorithm-based fault-tolerance (ABFT), which is weaving redundancy in the computation using error-correcting codes. In particular, we show that MatDot codes, an innovative code construction for parallel matrix multiplications, can be integrated into three-dimensional SUMMA (Scalable Universal Matrix Multiplication Algorithm [30]) in a communication-avoiding manner. To tolerate any two node failures, the proposed 3D Coded SUMMA requires ~50% less redundancy than replication, while the overhead in execution time is only about 5–10%.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Image Gradient Decomposition for Parallel and Memory-Efficient Ptychographic Reconstruction

Ptychography is a popular microscopic imaging modality for many scientific discoveries and sets the record for highest image resolution. Unfortunately, the high image resolution for ptychographic reconstruction requires significant amount of memory and computations, forcing many applications to compromise their image resolution in exchange for a smaller memory footprint and a shorter reconstruction time. In this paper, we propose a novel image gradient decomposition method that significantly reduces the memory footprint for ptychographic reconstruction by tessellating image gradients and diffraction measurements into tiles. In addition, we propose a parallel image gradient decomposition method that enables asynchronous point-to-point communications and parallel pipelining with minimal overhead on a large number of GPUs. Our experiments on a Titanate material dataset (PbTiO3) with 16632 probe locations show that our Gradient Decomposition algorithm reduces memory footprint by 51 times. In addition, it achieves time-to-solution within 2.2 minutes by scaling to 4158 GPUs with a super-linear strong scaling efficiency at 364% compared to runtimes at 6 GPUs. This performance is 2.7 times more memory efficient, 9 times more scalable and 86 times faster than the state-of-the-art algorithm.

Wang, Xiao↗
DOE OSTI2020

SQUARE: Strategic Quantum Ancilla Reuse for Modular Quantum Programs via Cost-Effective Uncomputation

Compiling high-level quantum programs to machines that are size constrained (i.e. limited number of quantum bits) and time constrained (i.e. limited number of quantum operations) is challenging. In this paper, we present SQUARE (Strategic QUantum Ancilla REuse), a compilation infrastructure that tackles allocation and reclamation of scratch qubits (called ancilla) in modular quantum programs. At its core, SQUARE strategically performs uncomputation to create opportunities for qubit reuse. Current Noisy Intermediate-Scale Quantum (NISQ) computers and forward-looking Fault-Tolerant (FT) quantum computers have fundamentally different constraints such as data locality, instruction parallelism, and communication overhead. Our heuristic-based ancilla-reuse algorithm balances these considerations and fits computations into resource-constrained NISQ or FT quantum machines, throttling parallelism when necessary. To precisely capture the workload of a program, we propose an improved metric, the "active quantum volume," and use this metric to evaluate the effectiveness of our algorithm. Furthermore, our results show that SQUARE improves the average success rate of NISQ applications by 1.47X. Surprisingly, the additional gates for uncomputation create ancilla with better locality, and result in substantially fewer swap gates and less gate noise overall. SQUARE also achieves an average reduction of 1.5X (and up to 9.6X) in active quantum volume for FT machines.

compiler optimization↗
DOE OSTI2020

Distributed Berkeley Efficient Long-Read to Long-Read Aligner and Overlapper (DiBELLA) v1.0.0

We present a parallel algorithm and scalable implementation for genome analysis, specifically the problem of finding overlaps and alignments for data from "third generation" long read sequencers. While long sequences of DNA offer enormous advantages for biological analysis and insight, current long read sequencing instruments have high error rates and therefore require different approaches to analysis than their short read counterparts. Our work focuses on an efficient distributed-memory parallelization of an accurate single-node algorithm for overlapping and aligning long reads. We achieve scalability of this irregular algorithm by addressing the competing issues of increasing parallelism, minimizing communication, constraining the memory footprint, and ensuring good load balance. The resulting application, DiBELLA, is the first distributed memory overlapper and aligner specifically designed for long reads and parallel scalability.

Ellis, Marquita↗
DOE OSTI2021

Online data analysis and reduction: An important co-design motif for extreme-scale computers

A growing disparity between supercomputer computation speeds and I/O rates means that it is rapidly becoming infeasible to analyze supercomputer application output only after that output has been written to a file system. Instead, data-generating applications must run concurrently with data reduction and/or analysis operations, with which they exchange information via high-speed methods such as interprocess communications. The resulting parallel computing motif, online data analysis and reduction (ODAR), has important implications for both application and HPC systems design. Here we introduce the ODAR motif and its co-design concerns, describe a co-design process for identifying and addressing those concerns, present tools that assist in the co-design process, and present case studies to illustrate the use of the process and tools in practical settings.

Data Analysis↗
DOE OSTI2021

Asynchronous distributed-memory task-parallel algorithm for compressible flows on unstructured 3D Eulerian grids

Here, we discuss the implementation of a finite element method, used to numerically solve the Euler equations of compressible flows, using an asynchronous runtime system (RTS). The algorithm is implemented for distributed-memory machines, using stationary unstructured 3D meshes, combining data-, and task-parallelism on top of the Charm++ RTS. Charm++’s execution model is asynchronous by default, allowing arbitrary overlap of computation and communication. Task-parallelism allows scheduling parts of an algorithm independently of, or dependent on, each other. Built-in automatic load balancing enables continuous redistribution of computational load by migration of work units based on real-time CPU load measurement. The RTS also features automatic checkpointing, fault tolerance, resilience against hardware failure, and supports power-, and energy-aware computation. We demonstrate scalability up to 25 x 10 9 cells at $\mathscr{O}$10 4 compute cores and the benefits of automatic load balancing for irregular workloads. The full source code with documentation is available at https://quinoacomputing.org.

42 ENGINEERING↗
DOE OSTI2024

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↗
DOE OSTI2020

Scalable All-pairs Shortest Paths for Huge Graphs on Multi-GPU Clusters

We present an optimized Floyd-Warshall (Floyd-Warshall) algorithm that computes the All-pairs shortest path (APSP) for GPU accelerated clusters. The Floyd-Warshall algorithm due to its structural similarities to matrix-multiplication is well suited for highly parallel GPU architectures. To achieve high parallel efficiency, we address two key algorithmic challenges: reducing high communication overhead and addressing limited GPU memory. To reduce high communication costs, we redesign the parallel (a) to expose more parallelism, (b) aggressively overlap communication and computation with pipelined and asynchronous scheduling of operations, and (c) tailored MPI-collective. To cope with limited GPU memory, we employ an offload model, where the data resides on the host and is transferred to GPU on-demand. The proposed optimizations are supported with detailed performance models for tuning. Our optimized parallel Floyd-Warshall implementation is up to 5x faster than a strong baseline and achieves 8.1 PetaFLOPS/sec on 256~nodes of the Summit supercomputer at Oak Ridge National Laboratory. This performance represents 70% of the theoretical peak and 80% parallel efficiency. The offload algorithm can handle 2.5x larger graphs with a 20% increase in overall running time.

Sao, Piyush↗
DOE OSTI2021

Devastator Parallel Discrete Event Simulation Runtime (Devastator) v1.0

The Devastator runtime is a modern C++ implementation of optimistic parallel discrete event simulation methods. Devastator allows simulation application code to productively specify their component and event functionality with C++14 constructs. It utilizes GASNet-EX for distributed memory communication and includes parallel performance optimizations such as light-weight thread message queues and asynchronous GVT. Furthermore, it supports efficient event broadcasts and pause-rewind-resume functionality to support periodic load balancing and outer loop optimization algorithms.

Chan, Cy↗
DOE OSTI2024

Enhanced quantum state transfer by circumventing quantum chaotic behavior

The ability to realize high-fidelity quantum communication is one of the many facets required to build generic quantum computing devices. In addition to quantum processing, sensing, and storage, transferring the resulting quantum states demands a careful design that finds no parallel in classical communication. Existing experimental demonstrations of quantum information transfer in solid-state quantum systems are largely confined to small chains with few qubits, often relying upon non-generic schemes. Here, by using a superconducting quantum circuit featuring thirty-six tunable qubits, accompanied by general optimization procedures deeply rooted in overcoming quantum chaotic behavior, we demonstrate a scalable protocol for transferring few-particle quantum states in a two-dimensional quantum network. These include single-qubit excitation, two-qubit entangled states, and two excitations for which many-body effects are present. Our approach, combined with the quantum circuit’s versatility, paves the way to short-distance quantum communication for connecting distributed quantum processors or registers, even if hampered by inherent imperfections in actual quantum devices.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗