arXiv subjects

Amanda Bienz

Publications and source records attributed to Amanda Bienz.

17 recordsLinked to original sources

arXiv2026

The Case of the Mysterious Citations

Mysterious citations are routinely appearing in peer-reviewed publications throughout the scientific community. In this paper, we developed an automated pipeline and examine the proceedings of four major high-performance computing conferences, comparing the accuracy of citations between the 2021 and 2025 proceedings. While none of the 2021 papers contained mysterious citations, every 2025 proceeding did, impacting 2-6% of published papers. In addition, we observe a sharp rise in paper title and authorship errors, motivating the need for stronger citation-verification practice. No author within our dataset acknowledged using AI to generate citations even though all four conference policies required it, indicating current policies are insufficient.

cs.DL↗

arXiv2026

Scaling All-to-all Operations Across Emerging Many-Core Supercomputers

Performant all-to-all collective operations in MPI are critical to fast Fourier transforms, transposition, and machine learning applications. There are many existing implementations for all-to-all exchanges on emerging systems, with the achieved performance dependent on many factors, including message size, process count, architecture, and parallel system partition. This paper presents novel all-to-all algorithms for emerging many-core systems. Further, the paper presents a performance analysis against existing algorithms and system MPI, with novel algorithms achieving up to 3x speedup over system MPI at 32 nodes of state-of-the-art Sapphire Rapids systems.

cs.DC↗

arXiv2025

Persistent and Partitioned MPI for Stencil Communication

Many parallel applications rely on iterative stencil operations, whose performance are dominated by communication costs at large scales. Several MPI optimizations, such as persistent and partitioned communication, reduce overheads and improve communication efficiency through amortized setup costs and reduced synchronization of threaded sends. This paper presents the performance of stencil communication in the Comb benchmarking suite when using non blocking, persistent, and partitioned communication routines. The impact of each optimization is analyzed at various scales. Further, the paper presents an analysis of the impact of process count, thread count, and message size on partitioned communication routines. Measured timings show that persistent MPI communication can provide a speedup of up to 37% over the baseline MPI communication, and partitioned MPI communication can provide a speedup of up to 68%.

cs.DC↗

arXiv2025

Optimizing Allreduce Operations for Modern Heterogeneous Architectures with Multiple Processes per GPU

Large inter-GPU all-reduce operations, prevalent throughout deep learning, are bottlenecked by communication costs. Emerging heterogeneous architectures are comprised of complex nodes, often containing $4$ GPUs and dozens to hundreds of CPU cores per node. Parallel applications are typically accelerated on the available GPUs, using only a single CPU core per GPU while the remaining cores sit idle. This paper presents novel optimizations to large GPU-aware all-reduce operations by extending the lane-aware algorithm to heterogeneous architectures and notably using multiple CPU cores per GPU to accelerate these operations. Using GPUDirect RDMA and host copy communications respectively, these multi-CPU-accelerated GPU-aware all-reduces yield speedups over system MPI of up to $3$x on LLNL's Tuolumne supercomputer and up to $2.45$x for large MPI all-reduces across the NVIDIA A100 GPUs of NCSA's Delta supercomputer.

cs.DC↗

arXiv2023

MPI Advance : Open-Source Message Passing Optimizations

The large variety of production implementations of the message passing interface (MPI) each provide unique and varying underlying algorithms. Each emerging supercomputer supports one or a small number of system MPI installations, tuned for the given architecture. Performance varies with MPI version, but application programmers are typically unable to achieve optimal performance with local MPI installations and therefore rely on whichever implementation is provided as a system install. This paper presents MPI Advance, a collection of libraries that sit on top of MPI, optimizing the underlying performance of any existing MPI library. The libraries provide optimizations for collectives, neighborhood collectives, partitioned communication, and GPU-aware communication.

cs.DC↗

arXiv2023

A More Scalable Sparse Dynamic Data Exchange

Parallel architectures are continually increasing in performance and scale, while underlying algorithmic infrastructure often fail to take full advantage of available compute power. Within the context of MPI, irregular communication patterns create bottlenecks in parallel applications. One common bottleneck is the sparse dynamic data exchange, often required when forming communication patterns within applications. There are a large variety of approaches for these dynamic exchanges, with optimizations implemented directly in parallel applications. This paper proposes a novel API within an MPI extension library, allowing for applications to utilize the variety of provided optimizations for sparse dynamic data exchange methods. Further, the paper presents novel locality-aware sparse dynamic data exchange algorithms. Finally, performance results show significant speedups up to 20x with the novel locality-aware algorithms.

cs.DC↗

arXiv2023

Collective-Optimized FFTs

This paper measures the impact of the various alltoallv methods. Results are analyzed within Beatnik, a Z-model solver that is bottlenecked by HeFFTe and representative of applications that rely on FFTs.

cs.MS↗

arXiv2023

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 messages using point-to-point communications, and any optimizations are added directly into the application. As a result, these optimizations lack portability. There is no easy way 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 optimizing 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 an up to 1.32x speedup on sparse matrix-vector multiplication within a BoomerAMG solve through the use of our optimized neighbor collectives. The authors analyze multiple implementations of neighborhood collectives, including a standard implementation, which simply wraps standard point-to-point communication, as well as multiple implementations of locality-aware aggregation. 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.

cs.DC↗

arXiv2022

Characterizing the Performance of Node-Aware Strategies for Irregular Point-to-Point Communication on Heterogeneous Architectures

Supercomputer architectures are trending toward higher computational throughput due to the inclusion of heterogeneous compute nodes. These multi-GPU nodes increase on-node computational efficiency, while also increasing the amount of data to be communicated and the number of potential data flow paths. In this work, we characterize the performance of irregular point-to-point communication with MPI on heterogeneous compute environments through performance modeling, demonstrating the limitations of standard communication strategies for both device-aware and staging-through-host communication techniques. Presented models suggest staging communicated data through host processes then using node-aware communication strategies for high inter-node message counts. Notably, the models also predict that node-aware communication utilizing all available CPU cores to communicate inter-node data leads to the most performant strategy when communicating with a high number of nodes. Model validation is provided via a case study of irregular point-to-point communication patterns in distributed sparse matrix-vector products. Importantly, we include a discussion on the implications model predictions have on communication strategy design for emerging supercomputer architectures.

cs.DC↗

arXiv2022

A Locality-Aware Bruck Allgather

Collective algorithms are an essential part of MPI, allowing application programmers to utilize underlying optimizations of common distributed operations. The MPI_Allgather gathers data, which is originally distributed across all processes, so that all data is available to each process. For small data sizes, the Bruck algorithm is commonly implemented to minimize the maximum number of messages communicated by any process. However, the cost of each step of communication is dependent upon the relative locations of source and destination processes, with non-local messages, such as inter-node, significantly more costly than local messages, such as intra-node. This paper optimizes the Bruck algorithm with locality-awareness, minimizing the number and size of non-local messages to improve performance and scalability of the allgather operation

cs.DC↗

arXiv2022

Performance Analysis and Optimal Node-Aware Communication for Enlarged Conjugate Gradient Methods

Krylov methods are a key way of solving large sparse linear systems of equations, but suffer from poor strong scalabilty on distributed memory machines. This is due to high synchronization costs from large numbers of collective communication calls alongside a low computational workload. Enlarged Krylov methods address this issue by decreasing the total iterations to convergence, an artifact of splitting the initial residual and resulting in operations on block vectors. In this paper, we present a performance study of an Enlarged Krylov Method, Enlarged Conjugate Gradients (ECG), noting the impact of block vectors on parallel performance at scale. Most notably, we observe the increased overhead of point-to-point communication as a result of denser messages in the sparse matrix-block vector multiplication kernel. Additionally, we present models to analyze expected performance of ECG, as well as, motivate design decisions. Most importantly, we introduce a new point-to-point communication approach based on node-aware communication techniques that increases efficiency of the method at scale.

cs.DC↗

arXiv2020

Modeling Data Movement Performance on Heterogeneous Architectures

The cost of data movement on parallel systems varies greatly with machine architecture, job partition, and nearby jobs. Performance models that accurately capture the cost of data movement provide a tool for analysis, allowing for communication bottlenecks to be pinpointed. Modern heterogeneous architectures yield increased variance in data movement as there are a number of viable paths for inter-GPU communication. In this paper, we present performance models for the various paths of inter-node communication on modern heterogeneous architectures, including the trade-off between GPUDirect communication and copying to CPUs. Furthermore, we present a novel optimization for inter-node communication based on these models, utilizing all available CPU cores per node. Finally, we show associated performance improvements for MPI collective operations.

cs.DC↗

arXiv2019

Node-Aware Improvements to Allreduce

The \texttt{MPI\_Allreduce} collective operation is a core kernel of many parallel codebases, particularly for reductions over a single value per process. The commonly used allreduce recursive-doubling algorithm obtains the lower bound message count, yielding optimality for small reduction sizes based on node-agnostic performance models. However, this algorithm yields duplicate messages between sets of nodes. Node-aware optimizations in MPICH remove duplicate messages through use of a single master process per node, yielding a large number of inactive processes at each inter-node step. In this paper, we present an algorithm that uses the multiple processes available per node to reduce the maximum number of inter-node messages communicated by a single process, improving the performance of allreduce operations, particularly for small message sizes.

cs.DC↗

arXiv2019

Reducing Communication in Algebraic Multigrid with Multi-step Node Aware Communication

Algebraic multigrid (AMG) is often viewed as a scalable $\mathcal{O}(n)$ solver for sparse linear systems. Yet, parallel AMG lacks scalability due to increasingly large costs associated with communication, both in the initial construction of a multigrid hierarchy as well as the iterative solve phase. This work introduces a parallel implementation of AMG to reduce the cost of communication, yielding an increase in scalability. Standard inter-process communication consists of sending data regardless of the send and receive process locations. Performance tests show notable differences in the cost of intra- and inter-node communication, motivating a restructuring of communication. In this case, the communication schedule takes advantage of the less costly intra-node communication, reducing both the number and size of inter-node messages. Node-centric communication extends to the range of components in both the setup and solve phase of AMG, yielding an increase in the weak and strong scalability of the entire method.

cs.DC↗

arXiv2018

Improving Performance Models for Irregular Point-to-Point Communication

Parallel applications are often unable to take full advantage of emerging parallel architectures due to scaling limitations, which arise due to inter-process communication. Performance models are used to analyze the sources of communication costs. However, traditional models for point-to-point communication fail to capture the full cost of many irregular operations, such as sparse matrix methods. In this paper, a node-aware based model is presented. Furthermore, the model is extended to include communication queue search time as well as an additional parameter estimating network contention. The resulting model is applied to a variety of irregular communication patterns throughout matrix operations, displaying improved accuracy over traditional models.

cs.DC↗

arXiv2016

Node Aware Sparse Matrix-Vector Multiplication

The sparse matrix-vector multiply (SpMV) operation is a key computational kernel in many simulations and linear solvers. The large communication requirements associated with a reference implementation of a parallel SpMV result in poor parallel scalability. The cost of communication depends on the physical locations of the send and receive processes: messages injected into the network are more costly than messages sent between processes on the same node. In this paper, a node aware parallel SpMV (NAPSpMV) is introduced to exploit knowledge of the system topology, specifically the node-processor layout, to reduce costs associated with communication. The values of the input vector are redistributed to minimize both the number and the size of messages that are injected into the network during a SpMV, leading to a reduction in communication costs. A variety of computational experiments that highlight the efficiency of this approach are presented.

cs.DC↗

arXiv2015

Reducing Parallel Communication in Algebraic Multigrid through Sparsification

Algebraic multigrid (AMG) is an $\mathcal{O}(n)$ solution process for many large sparse linear systems. A hierarchy of progressively coarser grids is constructed that utilize complementary relaxation and interpolation operators. High-energy error is reduced by relaxation, while low-energy error is mapped to coarse-grids and reduced there. However, large parallel communication costs often limit parallel scalability. As the multigrid hierarchy is formed, each coarse matrix is formed through a triple matrix product. The resulting coarse-grids often have significantly more nonzeros per row than the original fine-grid operator, thereby generating high parallel communication costs on coarse-levels. In this paper, we introduce a method that systematically removes entries in coarse-grid matrices after the hierarchy is formed, leading to an improved communication costs. We sparsify by removing weakly connected or unimportant entries in the matrix, leading to improved solve time. The main trade-off is that if the heuristic identifying unimportant entries is used too aggressively, then AMG convergence can suffer. To counteract this, the original hierarchy is retained, allowing entries to be reintroduced into the solver hierarchy if convergence is too slow. This enables a balance between communication cost and convergence, as necessary. In this paper we present new algorithms for reducing communication and present a number of computational experiments in support.

cs.DC↗