SEARCH · Engineering Papers

Results for “access pattern”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 91 records · Page 5

All sources NASA NTRS DOE OSTI

DOE OSTI2024

Distributed Multi-GPU Community Detection on Exascale Computing Platforms

Community detection is a fundamental operation in graph mining, and by uncovering hidden structures and patterns within complex systems it helps solve fundamental problems pertaining to social networks, such as information diffusion, epidemics, and recommender systems. Scaling graph algorithms for massive networks becomes challenging on modern distributed-memory multi-GPU (Graphics Processing Unit) systems due to limitations such as irregular memory access patterns, load imbalances, higher communication-computation ratios, and cross-platform support. We present a novel algorithm HiPDPL-GPU (distributed parallel Louvain) to address these challenges. We conduct experiments involving different partitioning techniques to achieve optimized performance of HiPDPL-GPU on the two largest supercomputers: Frontier and Summit. Remarkably, HiPDPL-GPU processes a graph with 4.2 billion edges in less than 3 minutes using 1024 GPUs. Qualitatively performance of HiPDPL-GPU is similar or better compared to other state-of-the-art CPU- and GPU-based implementations. While prior GPU implementations have predominantly employed CUDA, our first-of-its-kind implementation for community detection is cross-platform, accommodating both AMD and NVIDIA GPUs.

graph algorithms, high performance comptuing↗

DOE OSTI2024

Distributed Multi-GPU Community Detection on Exascale Computing Platforms

Community detection is a fundamental operation in graph mining, and by uncovering hidden structures and patterns within complex systems it helps solve fundamental problems pertaining to social networks, such as information diffusion, epidemics, and recommender systems. Scaling graph algorithms for massive networks becomes challenging on modern distributed-memory multi-GPU (Graphics Processing Unit) systems due to limitations such as irregular memory access patterns, load imbalances, higher communication-computation ratios, and cross-platform support. We present a novel algorithm HiPDPL-GPU (Distributed Parallel Louvain) to address these challenges. We conduct experiments involving different partitioning techniques to achieve an optimized performance of HiPDPL-GPU on the two largest supercomputers: Frontier and Summit. Remarkably, HiPDPL-GPU processes a graph with 4.2 billion edges in less than 3 minutes using 1024 GPUs. Qualitatively, the performance of HiPDPL-GPU is similar or better compared to other state-of-the-art CPU- and GPU-based implementations. While prior GPU implementations have predominantly employed CUDA, our first-of-its-kind implementation for community detection is cross-platform, accommodating both AMD and NVIDIA GPUs.

Sattar, Naw Safrin↗

DOE OSTI2023

Highly-scalable GPU-accelerated compressible reacting flow solver for modeling high-speed flows

Emerging supercomputing systems utilize a combination of central processing units (CPUs) and graphics processing units (GPUs) in an effort to reach exascale capabilities while minimizing the energy footprint of operating such systems. Such heterogeneous machines introduce new challenges for fluids solvers because the hardware architecture and operation of a GPU are fundamentally different from conventional CPUs. In this work, a general approach for efficient implementation of finite-volume based reacting flow solvers on such heterogeneous systems is presented. Three main challenges, namely, data access pattern, thread divergence, and thread safety, are addressed. Since compressible reacting flows require special methods to deal with chemical reactions, hyperbolic and nonlinear convection terms, and the presence of turbulence, specific algorithms that ensure GPU-based efficiency are developed. The approach is demonstrated on the widely available OpenFOAM open source software by modifying core algorithms for GPU accessibility. The scalability of the resulting solver, is demonstrated using practical test cases, including flow through a scramjet engine and the dynamics of a rotating detonation engine. Here, the solver provides near-ideal scaleup on a large number of GPUs (>3000), and extremely efficient use of the GPUs, with throughput nearly a constant even when processing a large number of control volumes.

42 ENGINEERING↗

DOE OSTI2016

Data Storage and sharing for the long tail of science

Research data infrastructure such as storage must now accommodate new requirements resulting from trends in research data management that require researchers to store their data for the long term and make it available to other researchers. We propose Data Depot, a system and service that provides capabilities for shared space within a group, shared applications, flexible access patterns and ease of transfer at Purdue University. We evaluate Depot as a solution for storing and sharing multiterabytes of data produced in the long tail of science with a use case in soundscape ecology studies from the Human- Environment Modeling and Analysis Laboratory. We observe that with the capabilities enabled by Data Depot, researchers can easily deploy fine-grained data access control, manage data transfer and sharing, as well as integrate their workflows into a High Performance Computing environment.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

DOE OSTI2019

MAC: Memory Access Coalescer for 3D-Stacked Memory

Emerging data-intensive applications, such as graph analytics and data mining, exhibit irregular memory access patterns. Research has shown that with these memory-bound applications, traditional cache-based processor architectures, which exploit locality and regular patterns to mitigate the memory-wall issue, are inefficient. Meantime, novel 3D-stacked memory devices, such as Hybrid Mem- ory Cube (HMC) and High Bandwidth Memory (HBM), promise significant increases in bandwidth that appear extremely appealing for memory-bound applications. However, conventional memory interfaces designed for cache-based architectures and JEDEC DDR devices fit poorly with the 3D-stacked memory, which leads to significant under-utilization of the promised high bandwidth. As a response to these issues, in this paper we propose MAC (Memory Access Coalescer), a coalescing unit for the 3D-stacked memory. We discuss the design and implementation of MAC, in the context of a custom designed cache-less architecture targeted at data-intensive, irregular applications. Through a custom simulation infrastructure based on the RISC-V toolchain, we show that MAC achieves a coalescing efficiency of 52.85% on average. It improves the performance of the memory system by 60.73% on average for a large set of irregular workloads.

Wang, Xi↗

DOE OSTI2013

Collective Memory Transfers for Multi-Core Chips

Future performance improvements for microprocessors have shifted from clock frequency scaling towards increases in on-chip parallelism. Performance improvements for a wide variety of parallel applications require domain-decomposition of data arrays from a contiguous arrangement in memory to a tiled layout for on-chip L1 data caches and scratchpads. How- ever, DRAM performance suffers under the non-streaming access patterns generated by many independent cores. We propose collective memory scheduling (CMS) that actively takes control of collective memory transfers such that requests arrive in a sequential and predictable fashion to the memory controller. CMS uses the hierarchically tiled arrays formal- ism to compactly express collective operations, which greatly improves programmability over conventional prefetch or list- DMA approaches. CMS reduces application execution time by up to 32% and DRAM read power by 2.2×, compared to a baseline DMA architecture such as STI Cell.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Memory access optimization for particle operations in computational fluid dynamics-discrete element method simulations

Computational Fluid Dynamics - Discrete Element Method is used to model gas-solid systems in several applications in energy, pharmaceutical and petrochemical industries. Computational performance bottlenecks often limit the problem sizes that can be simulated at industrial scale. The data structures used to store several millions of particles in such large-scale simulations have a large memory footprint that does not fit into the processor cache hierarchies on current high-performance-computing platforms, leading to reduced computational performance. This paper specifically addresses this aspect of memory access bottlenecks in industrial scale simulations. The use of space-filling curves to improve memory access patterns is described and their impact on computational performance is quantified in both shared and distributed memory parallelization paradigms. The Morton space filling curve applied to uniform grids and k-dimensional tree partitions are used to reorder the particle data-structure thus improving spatial and temporal locality in memory. The performance impact of these techniques when applied to two benchmark problems, namely the homogeneous-cooling-system and a fluidized-bed, are presented. We report these optimization techniques lead to approximately two-fold performance improvement in particle focused operations such as neighbor-list creation and data-exchange, with ~ 1.5 times overall improvement in a fluidization simulation with 1.27 million particles.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2015

Assessing a mini-application as a performance proxy for a finite element method engineering application

The performance of a large-scale, production-quality science and engineering application (‘app’) is often dominated by a small subset of the code. Even within that subset, computational and data access patterns are often repeated, so that an even smaller portion can represent the performance-impacting features. If application developers, parallel computing experts, and computer architects can together identify this representative subset and then develop a small mini-application (‘miniapp’) that can capture these primary performance characteristics, then this miniapp can be used to both improve the performance of the app as well as provide a tool for co-design for the high-performance computing community. However, a critical question is whether a miniapp can effectively capture key performance behavior of an app. This study provides a comparison of an implicit finite element semiconductor device modeling app on unstructured meshes with an implicit finite element miniapp on unstructured meshes. The goal is to assess whether the miniapp is predictive of the performance of the app. Finally, single compute node performance will be compared, as well as scaling up to 16,000 cores. Results indicate that the miniapp can be reasonably predictive of the performance characteristics of the app for a single iteration of the solver on a single compute node.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2021

Instruction Roofline: An insightful visual performance model for GPUs

The Roofline performance model provides an intuitive approach to identify performance bottlenecks and guide performance optimization. However, the classic FLOP-centric approach is inappropriate for the emerging applications that perform more integer operations than floating point operations. In this article, we reintroduce our Instruction Roofline Model on NVIDIA GPUs and expand our evaluation of it. The Instruction Roofline incorporates instructions and memory transactions across all memory hierarchies together, and provides more performance insights than the FLOP-oriented Roofline Model, that is, instruction throughput, stride memory access patterns, bank conflicts, and thread predication. We use our Instruction Roofline methodology to analyze eight proxy applications: HPGMG from AMReX, Matrix Transpose benchmarks, ADEPT from MetaHipMer's sequence alignment phase, EXTENSION from MetaHipMer's local assembly phase, CUSP, cuSPARSE, cudaTensorCoreGemm, and cuBLAS. We demonstrate the ability of our methodology to understand various aspects of performance and performance bottlenecks on NVIDIA GPUs and motivate code optimizations.

Ding, N↗

DOE OSTI2015

An empirical study of redundant array of independent solid-state drives (RAIS)

Solid-state drives (SSD) are popular storage media devices alongside magnetic hard disk drives (HDD). SSD flash chips are packaged in HDD form factors and SSDs are compatible with regular HDD device drivers and I/O buses. This compatibility allows easy replacement of individual HDDs with SSDs in existing storage systems. However, under certain circumstances, SSD write performance can be significantly slowed by garbage collection (GC) processes. The frequency of GC activity is directly correlated with the frequency of inside-SSD write operations and the amount of data written to it. GC scheduling is locally controlled by an internal SSD logic. Here, this paper studies the feasibility of Redundant Arrays of Independent Flash-based Solid-state drives (RAIS). We empirically analyze the RAIS performance using commercially-off-the-shelf (COTS) SSDs. We investigate the performance of various RAIS configurations under a variety of I/O access patterns. Finally, we present our performance and cost comparisons of RAIS with a fast, PCIe-based COTS SSD, in terms of performance and cost.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2020

Mochi: Composing Data Services for High-Performance Computing Environments

Technology enhancements and the growing breadth of application workflows running on high-performance computing (HPC) platforms drive the development of new data services that provide high performance on these new platforms, provide capable and productive interfaces and abstractions for a variety of applications, and are readily adapted when new technologies are deployed. The Mochi framework enables composition of specialized distributed data services from a collection of connectable modules and subservices. Rather than forcing all applications to use a one-size-fits-all data staging and I/O software configuration, Mochi allows each application to use a data service specialized to its needs and access patterns. This paper introduces the Mochi framework and methodology. The Mochi core components and microservices are described. Examples of the application of the Mochi methodology to the development of four specialized services are detailed. Finally, a performance evaluation of a Mochi core component, a Mochi microservice, and a composed service providing an object model is performed. The paper concludes by positioning Mochi relative to related work in the HPC space and indicating directions for future work.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2017

An electrostatic Particle-In-Cell code on multi-block structured meshes

We present an electrostatic Particle-In-Cell (PIC) code on multi-block, locally structured, curvilinear meshes called Curvilinear PIC (CPIC). Multi-block meshes are essential to capture complex geometries accurately and with good mesh quality, something that would not be possible with single-block structured meshes that are often used in PIC and for which CPIC was initially developed. In spite of the structured nature of the individual blocks, multi-block meshes resemble unstructured meshes in a global sense and introduce several new challenges, such as the presence of discontinuities in the mesh properties and coordinate orientation changes across adjacent blocks, and polyjunction points where an arbitrary number of blocks meet. In CPIC, these challenges have been met by an approach that features: (1) a curvilinear formulation of the PIC method: each mesh block is mapped from the physical space, where the mesh is curvilinear and arbitrarily distorted, to the logical space, where the mesh is uniform and Cartesian on the unit cube; (2) a mimetic discretization of Poisson's equation suitable for multi-block meshes; and (3) a hybrid (logical-space position/physical-space velocity), asynchronous particle mover that mitigates the performance degradation created by the necessity to track particles as they move across blocks. The numerical accuracy of CPIC was verified using two standard plasma–material interaction tests, which demonstrate good agreement with the corresponding analytic solutions. And compared to PIC codes on unstructured meshes, which have also been used for their flexibility in handling complex geometries but whose performance suffers from issues associated with data locality and indirect data access patterns, PIC codes on multi-block structured meshes may offer the best compromise for capturing complex geometries while also maintaining solution accuracy and computational efficiency.

36 MATERIALS SCIENCE↗

DOE OSTI2017

Exploring performance and energy tradeoffs for irregular applications: A case study on the Tilera many-core architecture

High performance, parallel applications with irregular data accesses are becoming a critical workload class for modern systems. In particular, the execution of such workloads on emerging many-core systems is expected to be a significant component of applications in data mining, machine learning, scientific computing and graph analytics. However, power and energy constraints limit the capabilities of individual cores, memory hierarchy and on-chip interconnect of such systems, thus leading to architectural and software trade-os that must be understood in the context of the intended application’s behavior. Irregular applications are notoriously hard to optimize given their data-dependent access patterns, lack of structured locality and complex data structures and code patterns. We have ported two irregular applications, graph community detection using the Louvain method (Grappolo) and high-performance conjugate gradient (HPCCG), to the Tilera many-core system and have conducted a detailed study of platform-independent and platform-specific optimizations that improve their performance as well as reduce their overall energy consumption. To conduct this study, we employ an auto-tuning based approach that explores the optimization design space along three dimensions - memory layout schemes, GCC compiler flag choices and OpenMP loop scheduling options. We leverage MIT’s OpenTuner auto-tuning framework to explore and recommend energy optimal choices for different combinations of parameters. We then conduct an in-depth architectural characterization to understand the memory behavior of the selected workloads. Finally, we perform a correlation study to demonstrate the interplay between the hardware behavior and application characteristics. Using auto-tuning, we demonstrate whole-node energy savings and performance improvements of up to 49:6% and 60% relative to a baseline instantiation, and up to 31% and 45:4% relative to manually optimized variants.

auto tuning↗

DOE OSTI2016

I/O-aware bandwidth allocation for petascale computing systems

In the Big Data era, the gap between the storage performance and an appli- cation's I/O requirement is increasing. I/O congestion caused by concurrent storage accesses from multiple applications is inevitable and severely harms the performance. Conventional approaches either focus on optimizing an ap- plication's access pattern individually or handle I/O requests on a low-level storage layer without any knowledge from the upper-level applications. In this paper, we present a novel I/O-aware bandwidth allocation framework to coordinate ongoing I/O requests on petascale computing systems. The motivation behind this innovation is that the resource management system has a holistic view of both the system state and jobs' activities and can dy- namically control the jobs' status or allocate resource on the y during their execution. We treat a job's I/O requests as periodical subjobs within its lifecycle and transform the I/O congestion issue into a classical scheduling problem. Based on this model, we propose a bandwidth management mech- anism as an extension to the existing scheduling system. We design several bandwidth allocation policies with different optimization objectives either on user-oriented metrics or system performance. We conduct extensive trace- based simulations using real job traces and I/O traces from a production IBM Blue Gene/Q system at Argonne National Laboratory. Experimental results demonstrate that our new design can improve job performance by more than 30%, as well as increasing system performance.

I/O congestion↗

DOE OSTI2016

ASA-FTL: An adaptive separation aware flash translation layer for solid state drives

Here, the flash-memory based Solid State Drive (SSD) presents a promising storage solution for increasingly critical data-intensive applications due to its low latency (high throughput), high bandwidth, and low power consumption. Within an SSD, its Flash Translation Layer (FTL) is responsible for exposing the SSD’s flash memory storage to the computer system as a simple block device. The FTL design is one of the dominant factors determining an SSD’s lifespan and performance. To reduce the garbage collection overhead and deliver better performance, we propose a new, low-cost, adaptive separation-aware flash translation layer (ASA-FTL) that combines sampling, data clustering and selective caching of recency information to accurately identify and separate hot/cold data while incurring minimal overhead. We use sampling for light-weight identification of separation criteria, and our dedicated selective caching mechanism is designed to save the limited RAM resource in contemporary SSDs. Using simulations of ASA-FTL with both real-world and synthetic workloads, we have shown that our proposed approach reduces the garbage collection overhead by up to 28% and the overall response time by 15% compared to one of the most advanced existing FTLs. We find that the data clustering using a small sample size provides significant performance benefit while only incurring a very small computation and memory cost. In addition, our evaluation shows that ASA-FTL is able to adapt to the changes in the access pattern of workloads, which is a major advantage comparing to existing fixed data separation methods.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2017

Direct synthesis of all-inorganic heterostructured CdSe/CdS QDs in aqueous solution for improved photocatalytic hydrogen generation

Here we present a facile aqueous approach to synthesize heterostructured CdSe/CdS QDs with all-inorganic chalcogenide S2- ligands under mild conditions. High-resolution transmission electron microscopy (HRTEM), X-ray diffraction (XRD), Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), and steady-state emission spectroscopy demonstrate that the heterostructured CdSe/CdS QDs with sulfur-rich surface composition are formed by heterogeneous nucleation of Cd2+ and S2- precursors on the CdSe QDs. After adsorption of small Ni(OH)(2) clusters over the surface in situ, the CdSe/CdS-Ni(OH)(2) photocatalyst enables H-2 production efficiently with an internal quantum yield of 52% under visible light irradiation at 455 nm, up to an 8-fold increase of activity to that of spherical CdSe QDs-Ni(OH)(2) under the same conditions. Femtosecond transient absorption spectroscopy, X-ray transient absorption (XTA) spectroscopy, steady-state and time-resolved emission spectroscopy show that the quasi-type-II band alignment in the CdSe/CdS heterostructure is responsible for the efficiency enhancement of light harvesting and surface/interfacial charge separation in solar energy conversion. The unprecedented results exemplify an easily accessible pattern of aqueous synthesis of all-inorganic heterostructured QDs for advanced photosynthetic H-2 evolution.

Li, Zhi-Jun↗

DOE OSTI2019

Concurrent data structures in the ATLAS offline software

In preparation for Run 3 of the LHC, the ATLAS experiment is modifying its offline software to be fully multithreaded. An important part of this is data structures that can be concurrently accessed from many threads both efficiently and safely. A standard way of achieving this is through mutual exclusion; however, the overhead from this can sometimes be excessive. Fully lockless implementations are known for some data structures; however, they are typically complex, and the overhead they require can sometimes be larger than that required for locking implementations. An interesting compromise is to allow lockless access only for reading but not for writing. This often allows the data structures to be much simpler, while still giving good performance for read-mostly access patterns. This proceeding shows some examples of this strategy in data structures used by the ATLAS offline software. It also gives examples of synchronization strategies inspired by read-copy-update, as well as helpers for memorizing values in a multithreaded environment.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2024

Boosting RDataFrame performance with transparent bulk event processing

RDataFrame is ROOT’s high-level interface for Python and C++ data analysis. Since it first became available, RDataFrame adoption has grown steadily and it is now poised to be a major component of analysis software pipelines for LHC Run 3 and beyond. Thanks to its design inspired by declarative programming principles, RDataFrame enables the development of highperformance, highly parallel analyses without requiring expert knowledge of multi-threading and I/O: user logic is expressed in terms of self-contained, small computation kernels tied together by a high-level API. This design completely decouples analysis logic from its actual execution, and opens several interesting avenues for workflow optimization. In particular, in this work we explore the benefits of moving internal data processing from an event-by-event to a bulkby-bulk loop. This refactoring dramatically reduces the framework’s runtime overheads; in collaboration with the I/O layer it improves data access patterns; it exposes information that optimizing compilers might use to auto-vectorize the invocation of user-defined computations; finally, while existing user-facing interfaces remain unaffected, it becomes possible to additionally offer interfaces that explicitly expose bulks of events, useful e.g. for the injection of GPU kernels into the analysis workflow. In order to inform similar future R&D, design challenges will be presented, as well as an investigation of the relevant timememory trade-off backed by novel performance benchmarks.

Guiraud, Enrico↗