检索结果-内蒙古大学图书馆

Unified programming models for Heterogeneous High-Performance Computers

Journal of Computer Science & Technology 2023年第1期38卷 211-218页

作者：马子轩金煜阳唐适之王豪杰薛伟诚翟季冬郑纬民 Department of Computer Science and Technology Tsinghua UniversityBeijing 100084China

Unified programming models can effectively improve program portability on various heterogeneous high-performance *** unified programming models put a lot of effort to code portability but are still far from achieving good performance *** this paper,we present a preliminary design of a performance-portable unified programming model including four aspects:programming language,programming abstraction,compilation optimization,and scheduling ***,domain-specific languages introduce domain knowledge to decouple the optimizations for different applications and *** unified programming abstraction unifies the common features of different architectures to support common ***-level compilation optimization enables comprehensive performance optimization based on multi-level intermediate ***-aware lightweight runtime scheduling system improves the resource utilization of heterogeneous *** is a perspective paper to show our viewpoints on programming models for emerging heterogeneous systems.

关键词： performance portability programming model heterogeneous supercomputer

来源：评论

学校读者我要写书评

暂无评论

ISP Agent: A Generalized In-storage-processingWorkload Offloading Framework by Providing Multiple Optimization Opportunities

引用

ACM TRANSACTIONS ON ARCHITECTURE AND CODE OPTIMIZATION 2024年第1期21卷 1-24页

作者： Kang, Seokwon Kim, Jongbin Lee, Gyeongyong Lee, Jeongmyung Seo, Jiwon Jung, Hyungsoo Song, Yong Ho Park, Yongjun Hanyang Univ Seoul South Korea Samsung Elect Suwon South Korea Yonsei Univ Seoul South Korea Hanyang Univ 220 Wangsimni Ro Seoul South Korea Samsung Elect 416 Maetan 3 Dong Suwon Gyeonggi Do South Korea Yonsei Univ 50 Yonsei Ro Seoul South Korea

As solid-state drives (SSDs) with sufficient computing power have recently become the dominant devices in modern computer systems, in-storage processing (ISP), which processes data within the storage without transferring it to the host memory, is being utilized in various emerging applications. The main challenge of ISP is to deliver storage data to the offloaded workload. This is difficult because of the information gap between the host and storage, the data consistency problem between the host and offloaded workloads, and SSD-specific hardware limitations. Moreover, because the offloaded workloads use internal SSD resources, host I/O performance might be degraded due to resource conflicts. Although several ISP frameworks have been proposed, existing ISP approaches that do not deeply consider the internal SSD behavior are often insufficient to support efficient ISP workload offloading with high programmability. In this article, we propose an ISP agent, a lightweight ISP workload offloading framework for SSD devices. The ISP agent provides I/O and memory interfaces that allow users to run existing function codes on SSDs without major code modifications, and separates the resources for the offloaded workloads from the existing SSD firmware to minimize interference with host I/O processing. The ISP agent also provides further optimization opportunities for the offloaded workload by considering SSD architectures. We have implemented the ISP agent on the OpenSSD Cosmos+ board and evaluated its performance using synthetic benchmarks and a real-world ISP-assisted database checkpointing application. The experimental results demonstrate that the ISP agent enhances host application performance while increasing ISP programmability, and that the optimization opportunities provided by the ISP agent can significantly improve ISP-side performance without compromising host I/O processing.

关键词： In-storage processing solid state drive firmware programming model

来源：评论

学校读者我要写书评

暂无评论

Tensions between private and public benefit associated with carbon farming

引用

SUSTAINABLE PRODUCTION AND CONSUMPTION 2024年 49卷 387-397页

作者： Pham-Kieu, My Ives, Stephen Badgery, Warwick Harrison, Matthew Tom Univ Tasmania Tasmanian Inst Agr Launceston Tas 7248 Australia Univ Tasmania Coll Business & Econ Launceston Tas 7248 Australia New South Wales Dept Primary Ind Orange NSW 2800 Australia Vietnam Natl Univ Agr Hanoi Vietnam

Theory posits that land use sustainability could be improved through implementation of management regimes that balance land manager aspirations on the one hand with those of society on the other. Here, our aim was to contrast private and public benefits realised from grazing management approaches aimed at either maximising farm enterprise profit or soil organic carbon (SOC) accrual. When grazing management was conducted with the aim of maximising profit, a fast rotation, high stocking rate regime was optimal;enterprise profit was $AU55/ha/ year higher than optimal management approaches for maximising SOC accrual. In contrast, flexible grazing management (with varying duration based on leaf stage and biomass availability) tended maximise SOC accrual, with 0.4 t/ha/year more SOC accrual than the continuous grazing treatment. As concurrent adoption of multiple farm management regimes better reflect real-world practicalities, we also examined how optimal farm management differed when one or two grazing approaches were operationalised to maximise profit and SOC accrual simultaneously. Simultaneous adoption of multiple grazing treatments on farm elicited greater profit ($AU7/ha/ year higher than when one management approach adopted across the whole farm), with the most preferable regime being a combination of fast rotation with high stocking rate and flexible grazing. Effects of carbon price and farm size were more influential on production and profit compared with weather variability, even though the latter altered the trajectory of seasonal SOC accrual. For carbon prices of $AU31/t CO2-e, 2-e, flexible and high intensity fast rotational grazing were each adopted on around half of farm area;as carbon prices increased to $AU50/t CO2-e, 2-e, flexible grazing management across the entire farm resulted in the greatest profit and SOC (additional 0.21 t SOC/ha/year under the flexible grazing treatment cf. continuous grazing). Our results demonstrate that private land managemen

关键词： Livestock Trade-offs Co-benefits Rotational grazing programming model Sustainable production Greenhouse gas emissions Soil carbon Farm economics

来源：评论

学校读者我要写书评

暂无评论

SNCL: a supernode OpenCL implementation for hybrid computing arrays

引用

JOURNAL OF SUPERCOMPUTING 2024年第7期80卷 9471-9493页

作者： Tang, Tao Lu, Kai Peng, Lin Cui, Yingbo Fang, Jianbin Huang, Chun Wang, Ruibo Yang, Canqun Guo, Yifei Natl Univ Def Technol Coll Comp Sci & Technol Changsha 410073 Peoples R China

Heterogeneous computing has been developing continuously in the field of high-performance computing because of its high performance and energy efficiency. More and more accelerators have emerged, such as GPU, FPGA, DSP, AI accelerator, and so on. Usually, the accelerator is connected to the host CPU as a peripheral device to form a tightly coupled heterogeneous computing node, and then, a parallel system is constructed by multiple nodes. This organization is computationally efficient, but not flexible. When new accelerators appear, it is difficult to join the system that has been built. At the hardware level, we create an array of accelerators and connect them to the existing system through a high-speed network. At the software level, we dynamically organize computing resources from various arrays to build a virtual heterogeneous computing node. This approach also includes a standard programming environment. Therefore, it is a more flexible, elastic, and scalable heterogeneous computing organization. In this paper, a supernode OpenCL implementation is proposed for hybrid parallel computing systems, in which virtual supernodes can be dynamically constructed between different computing arrays, and a standard OpenCL environment is implemented based on RDMA communication of high-speed interconnection, which can be combined with the system-level MPI programming environment, thereby realizing the large-scale parallel computing of the hybrid array. SNCL is compatible with existing MPI/OpenCL programs without the need for additional modifications. Experiments show that the runtime overhead of the supernode OpenCL environment is very low, and it is suitable for deploying applications with high computing density and large data scale between different arrays to utilize their computing power without affecting scalability.

关键词： OpenCL Supernode High-performance computing programming model

来源：评论

学校读者我要写书评

暂无评论

DISC: A Domain-Interaction Based programming model With Support for Heterogeneous Execution 14

DISC: A Domain-Interaction Based Programming Model With Supp...

引用

International Conference on High Performance Computing, Networking, Storage and Analysis

作者： Kurt, Mehmet Can Agrawal, Gagan Ohio State Univ Dept Comp Sci & Engn Columbus OH 43210 USA

ISBN: (纸本)9781479955008

Several emerging trends are pointing to increasing heterogeneity among nodes and/or cores in HPC systems. Existing programming models, especially for distributed memory execution, typically have been designed to facilitate high performance on homogeneous systems. This paper describes a programming model and an associated runtime system we have developed to address the above need. The main concepts in the programming model are that of a domain and interactions between the domain elements. We explain how stencil computations, unstructured grid computations, and molecular dynamics applications can be expressed using these simple concepts. We show how inter-process communication can be handled efficiently at runtime just from the knowledge of domain interaction, for different types of applications. Subsequently, we develop techniques for the runtime system to automatically partition and re-partition the work among heterogeneous processors or nodes.

关键词： programming model heterogeneous support load balancing

来源：评论

学校读者我要写书评

暂无评论

A Compilation Flow for Parametric Dataflow: programming model, Scheduling, and Application to Heterogeneous MPSoC

A Compilation Flow for Parametric Dataflow: Programming Mode...

引用

International Conference on Compilers, Architecture and Synthesis for Embedded Systems (CASES)

作者： Dardaillon, Mickael Marquet, Kevin Risset, Tanguy Martin, Jerome Charles, Henri-Pierre Univ Lyon Inria INSA Lyon CITI Inria F-69621 Villeurbanne France Univ Grenoble Alpes F-38000 Grenoble France CEA Grenoble LETI F-38054 Grenoble France

ISBN: (纸本)9781450330503

Efficient programming of signal processing applications on embedded systems is a complex problem. High level models such as Synchronous dataflow (SDF) have been privileged candidates for dealing with this complexity. These models permit to express inherent application parallelism, as well as analysis for both verification and optimization. Parametric dataflow models aim at providing sufficient dynamicity to model new applications, while at the same time maintaining the high level of analyzability needed for efficient real life implementations. This paper presents a new compilation flow that targets parametric dataflows. Built on the LLVM compiler infrastructure, it offers an actor based C++ programming model to describe parametric graphs, a compilation front-end providing graph analysis features, and a retargetable back-end to map the application on real hardware. This paper gives an overview of this flow, with a specific focus on scheduling. The crucial gap between dataflow models and real hardware on which actor firing is not atomic, as well as the consequences on FIFOs sizing and execution pipelining are taken into account. The experimental results illustrate our compilation flow applied to compilation of 3GPP LTE-Advanced demodulation on a heterogeneous MPSoC with distributed scheduling features. This achieves performances similar to time-consuming hand made optimizations.

关键词： data flow programming model heterogeneous MPSoC compiler scheduling

来源：评论

学校读者我要写书评

暂无评论

Optimization consensus analysis for group decision making in view of non-transferable and transferable allocation schemes

引用

JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY 2024年第7期75卷 1325-1342页

作者： Meng, Fanyong Zhao, Dengyu Tan, Chunqiao Nanjing Univ Informat Sci & Technol Nanjing Peoples R China Nanjing Audit Univ 219 Ningliu Rd Nanjing 2410044 Jiangsu Peoples R China

Most practical decision-making problems commonly contain several alternatives evaluated by more than one decision maker (DM) according to the built criteria system. It is known as multi-criteria group decision making problems. The decision-making results should be based on highly similar individual opinions to derive the most suitable choice. Nevertheless, the divergence of individual judgments is usually significant due to various subjective and objective reasons. In this case, some DMs need to modify their judgments to derive final results, the consensus adjustment in group decision making. Like supply chain management, the centralized consensus process can minimize individual adjustments under the set consensus condition. Based on the hypothesis of rational humans that individuals always pursue maximum benefits, this paper proposes an iterative consensus process in a centralized environment. According to non-consensus adjustments, the iterative procedure contains non-transferable and transferable two parts. When the non-transferable consensus adjustment procedure is completed, the transferable consensus adjustment procedure is activated using an asymmetrical Nash bargaining game allocation mechanism. Additionally, a case study is provided to demonstrate the efficaciousness and practicality of the new method. Meanwhile, sensitivity and comparison analyses are done.

关键词： Group decision making non-transferable and transferable consensus adjustment programming model asymmetrical Nash bargaining game

来源：评论

学校读者我要写书评

暂无评论

BEELINE: BilevEl dEcomposition aLgorithm for synthesis of Industrial eNergy systEms

引用

COMPUTERS & CHEMICAL ENGINEERING 2024年 180卷

作者： Jimenez-Romero, Julia Azapagic, Adisa Smith, Robin Univ Manchester Ctr Proc Integrat Dept Chem Engn Manchester M13 9PL England Univ Manchester Dept Chem Engn Sustainable Ind Syst Grp Manchester M13 9PL England

Energy transition to a more sustainable basis is the most significant and complex challenge facing industry. On most industrial sites, the largest single energy user is the utility system that produces the heat and power necessary for the site. The heavy reliance of current utility systems on fossil fuels and the requirement of strategic measures to ensure a sustainable future has prompted researchers to explore different energy sources, technologies and pathways for evolving existing systems to a sustainable basis for future utility systems. Whether based on renewable or non-renewable energy sources, it is essential to minimize energy demand and mitigate emissions. The present work focuses on developing cost-effective solutions for the synthesis of process utility systems, considering site-wide energy integration. This will provide a platform in future work for the efficient introduction of renewable energy sources in the transition to sustainable systems. Utility system performance is generally determined by system configuration and operational load. Steam mains selection, in terms of pressures and superheating have a critical role in the utility system performance and site energy integration. Therefore, the synthesis of energy-efficient utility systems involves optimizing the configuration of utility components and number and operating conditions of steam mains simultaneously. Due to nonlinearities and non-convexities from underlying physics and binary decisions involved, the resulting Mixed-Integer Non-Linear Problem (MINLP) presents challenges for advanced state-of-art solvers to solve real-world problems. In past work, a number of important practical issues have been oversimplified in order to make the solution tractable. However, the oversimplifications also lead to misleading results. In this research, a mathematical formulation for simultaneous optimization of comprehensive utility system configurations and operating conditions is for the first time combine

关键词： Superstructure Nonconvex mixed integer nonlinear programming model Bilevel decomposition Mixed integer linear relaxation Heat recovery Process steam systems Steam networks

来源：评论

学校读者我要写书评

暂无评论

Consensus Reaching Process for Traditional Group Decision Making in View of the Optimal Adjustment Mechanism

引用

IEEE TRANSACTIONS ON CYBERNETICS 2023年第6期53卷 3748-3759页

作者： Meng, Fan-Yong Pedrycz, Witold Tang, Jie Nanjing Univ Informat Sci & Technol Sch Management Sci & Engn Nanjing 210044 Peoples R China Univ Alberta Dept Elect & Comp Engn Edmonton AB T6G 2R3 Canada

Consensus reaching process (CRP) is a key topic in the area of group decision making (GDM). When the consensus level is not high enough, it becomes necessary to adjust the original opinions of decision makers (DMs). To offer the adjustment reference for DMs, we build the programming models to determine the minimum modification to be carried out from the individual and global perspectives. Meanwhile, all DMs are divided into two subgroups: DMs with acceptable and unacceptable consensus levels. If some DMs with unacceptable consensus level do not accept the relevant modifications, the Nash bargaining game-based programming model is built for the fairness and efficiency of modifications. When some DMs refuse to make any modifications or tend to modify the opinions in their way, with respect to different group consensus situations, we make the minimum hybrid penalty mechanism by the Nash bargaining game-based programming models. For each case, we determine the corresponding optimal modification mechanism in view of the fixed individual total modification and the maximum consensus level. Furthermore, we study the arrangements of weights of DMs according to their cardinal and ordinal consensus contributions. Based on these results, we present a new algorithm and illustrate its application by a numerical example. Moreover, we carry out the sensitivity and comparison analysis. We summarize the conclusions and future research directions in the end. The main originality of the new method includes: the fairness and efficiency of modifications, and the determination of the hybrid penalty mechanism.

关键词： Costs Decision making Linguistics Social networking (online) programming profession Analytical models Optimization Consensus reaching process (CRP) group decision making (GDM) Nash bargaining game optimal modification mechanism programming model

来源：评论

学校读者我要写书评

暂无评论

Dedicated Instruction Set for Pattern-Based Data Transfers: An Experimental Validation on Systems Containing In-Memory Computing Units

引用

IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS 2023年第11期42卷 3757-3767页

作者： Mambu, Kevin Charles, Henri-Pierre Kooli, Maha Univ Grenoble Alpes CEA LIST F-38000 Grenoble France

In memory computing (IMC) aims at solving the performance gap between CPU and memories introduced by the memory wall. However, general-purpose IMC does not consider the optimization of data transfers for patterns, such as stencils and convolutions. This article proposes a new instruction set architecture (ISA) and a novel pattern encoding for IMC to transfer and organize data streams in order to perform efficiently computation. This instruction set is implemented on the data-locality management unit (DMU) as a subset of the computational SRAM (C-SRAM) ISA. A programming model to interact with the DMU at language level is also presented in this article. This DMU ISA is evaluated on six applications run on three different system nodes. These system nodes are based on existing RISC-V cores and range from embedded to high-performance computing domain. Experiments show on average a speed-up of x8.81, an energy reduction factor of x6.81, and an improvement of the number of operations per cycle of x4.59, for the C-SRAM architecture integrating the proposed ISA of the DMU compared to a reference implementation on embedded systems. Results also show an improvement of the number of operations per cycle of x2.99 compared to a reference implementation on all system nodes.

关键词： Computer architecture Instruction sets Random access memory programming Computational modeling Central Processing Unit Data transfer Convolution in-memory computing (IMC) instruction set architecture (ISA) non-von Neumann pattern performance analysis programming model stencil

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：