检索结果-内蒙古大学图书馆

ASHA: An adaptive shared-memory sharing architecture for multi-programmed GPUs

MICROPROCESSORS AND MICROSYSTEMS 2016年第PartB期46卷 264-273页

作者： Abbasitabar, Hamed Samavatian, Mohammad Hossein Sarbazi-Azad, Hamid Sharif Univ Technol Dept Comp Engn HPCAN Lab 434 Azadi Ave Tehran Iran

Spatial multi-programming is one of the most efficient multi-programming methods on Graphics Processing Units (GPUs). This multi-programming scheme generates variety in resource requirements of stream multiprocessors (SMs) and creates opportunities for sharing unused portions of each SM resource with other SMs. Although this approach drastically improves GPU performance, in some cases it leads to performance degradation due to the shortage of allocated resource to each program. Considering shared memory as one of the main bottlenecks of thread-level parallelism (TLP), in this paper, we propose an adaptive shared-memory sharing architecture, called ASHA. ASHA enhances spatial multi-programming performance and increases utilization of GPU resources. Experimental results demonstrate that ASHA improves speedup of a multi-programmed GPU by 17%-21%, on average, for 2- to 8-program execution scenarios, respectively. (C) 2016 Published by Elsevier B.V.

关键词： GPGPU multi-programming Resource sharing Shared-memory

来源：评论

学校读者我要写书评

暂无评论

Rinnegan: Efficient Resource Use in Heterogeneous Architectures 16

Rinnegan: Efficient Resource Use in Heterogeneous Architectu...

引用

International Conference on Parallel Architectures and Compilation (PACT)

作者： Panneerselvam, Sankaralingam Swift, Michael Univ Wisconsin Madison WI 53706 USA

ISBN: (纸本)9781450341219

Current processors provide a variety of different processing units to improve performance and power efficiency. For example, ARM's ***, AMD's APUs, and Oracle's M7 provide heterogeneous processors, on-die GPUs, and on-die accelerators. However, the performance experienced by programs using these processing units can vary widely due to contention from multiprogramming, thermal constraints and other issues. In these systems, the decision of where to execute a task must consider not only execution time of the task, but also current system conditions. We built Rinnegan, a Linux kernel extension and runtime library, to perform scheduling and handle task placement in heterogeneous systems. The Rinnegan kernel extension monitors and reports the utilization of all processing units to applications, which then makes placement decisions at user level. The Rinnegan runtime provides a performance model to predict the speedup and overhead of offloading a task. With this model and the current utilization of processing units, the runtime can select the task placement that best achieves an application's performance goals, such as low latency, high throughput, or real-time deadlines. When integrated with StarPU, a runtime system for heterogeneous architectures, Rinnegan improves StarPU by performing 1.5-2x better than its native scheduling policies in a shared heterogeneous environment.

关键词： operating systems multi-programming resource management accelerators task scheduling application adaptation heterogeneous architectures

来源：评论

学校读者我要写书评

暂无评论

Fairness Metrics for multi-Threaded Processors

IEEE COMPUTER ARCHITECTURE LETTERS

引用

IEEE COMPUTER ARCHITECTURE LETTERS 2011年第1期10卷 4-7页

作者： Vandierendonck, Hans Seznec, Andre Univ Ghent Dept Elect & Informat Syst Ghent Belgium INRIA Rennes Rennes France

multi-threaded processors execute multiple threads concurrently in order to increase overall throughput. It is well documented that multi-threading affects per-thread performance but, more importantly, some threads are affected more than others. This is especially troublesome for multi-programmed workloads. Fairness metrics measure whether all threads are affected equally. However defining equal treatment is not straightforward. Several fairness metrics for multi-threaded processors have been utilized in the literature, although there does not seem to be a consensus on what metric does the best job of measuring fairness. This paper reviews the prevalent fairness metrics and analyzes their main properties. Each metric strikes a different trade-off between fairness in the strict sense and throughput. We categorize the metrics with respect to this property. Based on experimental data for SMT processors, we suggest using the minimum fairness metric in order to balance fairness and throughput.

关键词： multi-threaded processors multi-programming measurement quality-of-service fairness

来源：评论

学校读者我要写书评

暂无评论

A hybrid open queuing network model approach for multi-threaded dataflow architecture

引用

COMPUTER COMMUNICATIONS 2008年第17期31卷 4098-4106页

作者： Bhaskar, Vidhyacharan Adjallah, Kondo Hloindo SRM Univ Dept Elect & Commun Engn Kancheepuram District 603203 Tamil Nadu India Univ Technol Troyes Charles Delaunay Inst F-10010 Troyes France

multi-threading has been proposed as an execution model for massively built parallel processors. Due to the large amount of potential parallelism, resource management is a critical issue in multi-threaded architecture. The challenge of multi-threading is to hide the latency by switching among a set of ready threads and thus to improve the processor utilization. Threads are dynamically scheduled to execute based on availability of data. In this paper, two hybrid open queuing network models are proposed. Two sets of processors: synchronization processors and execution processors exist. Each processor is modeled as a server serving a single-queue or multiple-servers serving a single-queue. Performance measures like response times, system throughput and average queue lengths are evaluated for both the hybrid models. The utilizations of the two models are derived and compared with each other. A mean value analysis is performed and different performance measures are plotted. Crown copyright (C) 2008 Published by Elsevier B.V. All rights reserved.

关键词： Synchronization and execution processors multi-programming Queue lengths Response times Throughput

来源：评论

学校读者我要写书评

暂无评论

A hybrid closed queuing network model for multi-threaded dataflow architecture

引用

COMPUTERS & ELECTRICAL ENGINEERING 2005年第8期31卷 556-571页

作者： Bhaskar, V Univ Technol Troyes Dept Gen Syst Infomat & Telecommun F-10010 Troyes France

In this paper, a closed queuing network model with both single and multiple servers has been proposed to model dataflow in a multi-threaded architecture. multi-threading is useful in reducing the latency by switching among a set of threads in order to improve the processor utilization. Two sets of processors, synchronization and execution processors exist. Synchronization processors handle load/store operations and execution processors handle arithmetic/logic and control operations. A closed queuing network model is suitable for large number of job arrivals. The normalization constant is derived using a recursive algorithm for the given model. State diagrams are drawn from the hybrid closed queuing network model, and the steady-state balance equations are derived from it. Performance measures such as average response times and average system throughput are derived and plotted against the total number of processors in the closed queuing network model. Other important performance measures like processor utilizations, average queue lengths, average waiting times and relative utilizations are also derived. (c) 2005 Elsevier Ltd. All rights reserved.

关键词： synchronization and execution processors multi-programming queue lengths response times utilizations normalization constant throughput

来源：评论

学校读者我要写书评

暂无评论

ABSENCE OF INDIVIDUAL STARVATION USING WEAK SEMAPHORES

引用

INFORMATION PROCESSING LETTERS 1986年第3期23卷 159-162页

作者： UDDING, JT Institute for Biomedical Computing Washington University St. Louis MO 63130 U.S.A.

As far as scheduling is concerned, there are 2 kinds of semaphores: weak and strong. When determining mutual exclusion problems, one typically assumes the existence of strong semaphores. A program is derived that demonstrates that strong semaphores can be implemented by weak ones. The techniques employed for the derivation are standard, and it is straightforward, with the arguments employed for the derivation, to deduce a formal correctness proof, which, for the purpose of this analysis, is considered superfluous. In this program, as in earlier programs that implement strong semaphores, the order of 2 V-operations -- V(enter) and V(queue) -- turns out to be critical. However, the reason is apparent. It merely stems from a mutual exclusion problem.

关键词： multi-programming operating systems analysis of algorithms

来源：评论

学校读者我要写书评

暂无评论

Essential Features of the CERN SPS Distributed Control System

引用

IFAC Proceedings Volumes 1981年第1期14卷 121-126页

作者： J. Altaber V. Frammery C. Gareyte P. van der Stok CERN European Organizaton for Nuclear Research Geneva Switzerland

The SPS accelerator presents a considerable industrial control problem with the additional complication that the control procedures are never fixed. Right from the beginning it was decided to base the control system on a distributed network making use of an interpretive language for the control processes. The success of these decisions can be seen from the fact that over the last six years, the system has grown to a network of more tha~ 50 computers spread over a ten square kilometer site, all the time controlling an ever-changing accelerator complex. This paper will discuss the major elements of the strategy used and explain the reason for their choice. Microprocessors have become very popular in the field of. industrial control and the SPS control system is going to integrate this trend with little difficulty. The paper will show that the SPS approach is ideally suited to the construction of a real-time control network making use only of microprocessor based units.

关键词： Distributed control system network multi-programming multi-processing

来源：评论

学校读者我要写书评

暂无评论

STUDY OF A PAGE-ON-DEMAND SYSTEM

引用

INFORMATION PROCESSING LETTERS 1977年第4期6卷 125-132页

作者： BRANDWAJN, A MOUNEIX, B ECOLE NATL SUPER TELECOMMUNICAT PARISFRANCE

来源：评论

学校读者我要写书评

暂无评论

GAME INTERPRETATION OF DEADLOCK AVOIDANCE PROBLEM

引用

COMMUNICATIONS OF THE ACM 1977年第10期20卷 741-745页

作者： DEVILLERS, R UNIV LIBRE BRUXELLES DEPT MATHINFORMAT THEOR LABB-1050 BRUSSELSBELGIUM

The deadlock avoidance problem may be defined informally as the determination, from some a priori information about the processes, resources, operating system, etc., of the 'safe situations' which may be realized without endangering the smooth running of the system. When each process specifies its future needs by a flowchart of need-defined steps, a global approach to the phenomenon and its interpretation as a game between the operating system and the processes allows formalization of risk and safety concepts. The bipartite graph representation of this game may then be used to construct explicitly the set of safe states and to study their properties. [ABSTRACT FROM AUTHOR]

关键词： dead-lock deadlock avoidance deadly embrace flowchart interlock multi-programming operating system resource allocation time-sharing

来源：评论

学校读者我要写书评

暂无评论

The Design of the Venus Operating System

引用

Communications of the ACM 1972年第3期15卷 144-149页

作者： Liskov, Barbara H. MITRE Corporation Bedford MA 01730 United States

The Venus Operating System is an experimental multiprogramming system which supports five or six concurrent users on a small computer. The system was produced to test the effect of machine architecture on complexity of software. The system is defined by a combination of microprograms and software. The microprogram defines a machine with some unusual architectural features;the software exploits these features to define the operating system as simply as possible. In this paper the development of the system is described, with particular emphasis on the principles which guided the design. © 1972, ACM. All rights reserved.

关键词： data sharing levels of abstraction machine architecture microprogramming multi-programming operating systems process communication processes resource management deadlock segments semaphores system design virtual devices virtual machines

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：