检索结果-内蒙古大学图书馆

10th International Symposium on Advanced Parallel Processing Technologies, APPT 2013

作者： Ye, Xiaochun Fan, Dongrui Wang, Da Song, Fenglong Zhang, Hao Tang, Zhimin State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing China

ISBN: (纸本)9783642452925

Fast multicore simulators are extremely useful in evaluating design alternatives and enabling early software development. Among the state-of-the-art multicore simulators, Simics is a very popular used one both in academia and industry. It has a powerful debugging system, and also provides an accelerator to support multithreaded or distributed simulation. However, this kind of parallel mechanism mainly aims at speed up distributed systems. It is not suitable for the shared-memory multicore systems which are much more commonly used. In this paper, we propose a novel parallel mechanism to improve the simulation speed of shared-memory multicore systems. More importantly, our approach is compatible with other optimizations and exist debugging systems used in Simics. Experimental results show that our parallel approach achieves an average speedup of 9.6x (up to 12.2x) when running SPLASH-2 kernel on a 16-core host machine. © 2013 Springer-Verlag.

关键词： Simulators

来源：评论

学校读者我要写书评

暂无评论

Exploiting the input sparsity to accelerate deep neural networks 24

Exploiting the input sparsity to accelerate deep neural netw...

引用

24th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2019

作者： Dong, Xiao Liu, Lei Li, Guangli Li, Jiansong Zhao, Peng Wang, Xueying Feng, Xiaobing State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences China State Key Laboratory of Computer Architecture University of Chinese Academy of Sciences China

ISBN: (纸本)9781450362252

Efficient inference of deep learning models are challenging and of great value in both academic and industrial community. In this paper, we focus on exploiting the sparsity in input data to improve the performance of deep learning models. We propose an end-to-end optimization pipeline to generate programs for the inference with sparse input. The optimization pipeline contains both domain-specific and general optimization techniques and is capable of generating eficient code without relying on the off-the-shelf libraries. Evaluations show that we achieve significant speedups over the state-of-the-art frameworks and libraries on a real-world application, e.g., 9.8× over TensorFlow and 3.6× over Intel MKL on the detection in autonomous driving. © 2019 Copyright held by the owner/author(s).

关键词： Libraries

来源：评论

学校读者我要写书评

暂无评论

Shared work list: Hacking amorphous data parallelism in UPC

Shared work list: Hacking amorphous data parallelism in UPC

引用

2012 International Workshop on Programming Models and Applications for Multicores and Manycores, PMAM 2012

作者： Xu, Shixiong Chen, Li State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781450312110

"Irregular" algorithms using data structures like sparse graphs, trees and sets prevail in the most emerging problems domains such as social network analysis, machine learning, data mining and computational science. The irregularity of underlying data structures leads to unstructured parallelism in these algorithms, consequently making it pretty hard for users to write efficient parallel implementations on distributed memory systems. Unified Parallel C language provides convenience of a global address space with the locality control needed for high performance and scalability. However, the Single Program Multiple Data execution model with a statically fixed set of executing threads makes UPC does not support applications with unstructured parallelism. In this paper, we first put forward Shared Work List to UPC and advocate a programming paradigm for writing applications with amorphous data parallelism on distributed memory systems. We also introduce user-assisted speculative execution based on Active Message model to support speculative execution on distributed memory systems. Efficient mechanism of work dispatching and related optimizations are presented as well. We preliminarily choose Breadth-first Search as a case study to demonstrate the feasibility, pro-grammability and performance benefits out of Shared Work List. © 2012 ACM.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

An optimized large-scale hybrid DGEMM design for CPUs and ATI GPUs 12

An optimized large-scale hybrid DGEMM design for CPUs and AT...

引用

26th ACM International Conference on Supercomputing, ICS'12

作者： Li, Jiajia Li, Xingjian Tan, Guangming Chen, Mingyu Sun, Ninghui State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781450313162

In heterogeneous systems that include CPUs and GPUs, the data transfers between these components play a critical role in determining the performance of applications. Software pipelining is a common approach to mitigate the overheads of those transfers. In this paper we investigate advanced software-pipelining optimizations for the double-precision general matrix multiplication (DGEMM) algorithm running on a heterogeneous system that includes ATI GPUs. Our approach decomposes the DGEMM workload to a finer detail and hides the latency of CPU-GPU data transfers to a higher degree than previous approaches in literature. We implement our approach in a five-stage software pipelined DGEMM and analyze its performance on a platform including x86 multi-core CPUs and an ATI Radeon™ HD5970 GPU that has two Cypress GPU chips on board. Our implementation delivers 758 GFLOPS (82% floating-point efficiency) when it uses only the GPU, and 844 GFLOPS (80% efficiency) when it distributes the workload on both CPU and GPU. We analyze the performance of our optimized DGEMM as the number of GPU chips employed grows from one to two, and the results show that resource contention on the PCIe bus and on the host memory are limiting factors. Copyright 2012 ACM.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

dCompaction： Speeding up Compaction of the LSM-Tree via Delayed Compaction

引用

Journal of computer Science & technology 2017年第1期32卷 41-54页

作者： Feng-Feng Pan Yin-Liang Yue Jin Xiong State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing 100190 China University of Chinese Academy of Sciences Beijing 100049 China Institute of Information Engineering Chinese Academy of Sciences Beijing 100093 China

key-value （KV） stores have become a backbone of large-scale applications in today＇s data centers. Write- optimized data structures like the Log-Structured Merge-tree （LSM-tree） and their variants are widely used in KV storage systems like BigTable and RocksDB. Conventional LSM-tree organizes KV items into multiple, successively larger components, and uses compaction to push KV items from one smaller component to another adjacent larger component until the KV items reach the largest component. Unfortunately, current compaction scheme incurs significant write amplification due to repeated KV item reads and writes, and then results in poor throughput. We propose a new compaction scheme, delayed compaction （dCompaction） that decreases write amplification, dCompaction postpones some compactions and gathers them into the following compaction. In this way, it avoids KV item reads and writes during compaction, and consequently improves the throughput of LSM-tree based KV stores. We implement dCompaction on RocksDB, and conduct extensive experiments. Validation using YCSB framework shows that compared with RocksDB, dCompaction has about 40% write performance improvements and also comparable read performance.

关键词： key-value store Log-Structured Merge-tree (LSM-tree) write amplification delayed compaction

来源：评论

学校读者我要写书评

暂无评论

Revisiting benchmarking principles and methodologies for big data benchmarking 6th

Revisiting benchmarking principles and methodologies for big...

引用

6th Workshop on Big Data Benchmarks, Performance Optimization, and Emerging Hardware, BPOE 2015

作者： Zhao, Liutao Gao, Wanling Jin, Yi Key Laboratory of Cloud Computing Technology and Application Beijing Computing Center Beijing China State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing China

ISBN: (纸本)9783319290058

Benchmarking as yardsticks for system design and evaluation, has developed a long period and plays a pivotal role in many domains, such as database systems and high performance computing. Through prolonged and unremitting efforts, benchmarks on these domains have been reaching their maturity gradually. However, in terms of emerging scenarios of big data, its different properties in data volume, data types, data processing requirements and techniques, make that existing benchmarks are rarely appropriate for big data systems and further make us wonder how to define a good big data benchmark. In this paper, we revisit successful benchmarks in other domains from two perspectives: benchmarking principles which define fundamental rules, and methodologies which guide the benchmark constructions. Further, we conclude the benchmarking principle and methodology on big data benchmarking from a recent open-source effort – BigDataBench. © Springer International Publishing Switzerland 2016.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

Constructing Edge-Colored Graph for Heterogeneous Networks

引用

Journal of computer Science & technology 2015年第5期30卷 1154-1160页

作者：侯睿武继刚陈亚文张海波隋秀峰 School of Computer Science and Software Engineering Tianjin Polytechnic University Tianjin 300387 China State Key Laboratory of Computer Architecture~ Institute of Computing Technology Chinese Academy of Sciences Beijing 100190 China Department of Computer Science University of Otago Dunedin 9054 New Zealand

In order to build a fault-tolerant network, heterogeneous facilities are arranged in the network to prevent homogeneous faults from causing serious damage. This paper uses edge-colored graph to investigate the features of a network topology which is survivable after a set of homogeneous devices malfunction. We propose an approach to designing such networks under arbitrary parameters. We also show that the proposed approach can be used to optimize inter-router connections in network-on-chip to reduce the additional consum!otion of energy and time delay.

关键词： network reliability homogeneous fault fault tolerance reconfigurable system network-on-chip

来源：评论

学校读者我要写书评

暂无评论

NfvInsight:A Framework for Automatically Deploying and Benchmarking VNF Chains

引用

Journal of computer Science & technology 2022年第3期37卷 680-698页

作者： Tian-Ni Xu Hai-Feng Sun Di Zhang Xiao-Ming Zhou Xiu-Feng Sui Sa Wang Qun Huang Yun-Gang Bao State Key Laboratory of Computer Architecture Institute of Computing TechnologyChinese Academy of SciencesBeijing 100190China University of Chinese Academy of Sciences Beijing 100049China School of Information and Electronics Beijing Institute of TechnologyBeijing 100081China Peng Cheng Laboratory Shenzhen 518055China Department of Computer Science and Technology Peking UniversityBeijing 100871China

With the advent of virtualization techniques and software-defined networking(SDN),network function virtualization(NFV)shifts network functions(NFs)from hardware implementations to software appliances,between which exists a performance *** to narrow the gap is an essential issue of current NFV ***,the cumbersomeness of deployment,the water pipe effect of virtual network function(VNF)chains,and the complexity of the system software stack together make it tough to figure out the cause of low performance in the NFV *** pinpoint the NFV system performance,we propose NfvInsight,a framework for automatic deployment and benchmarking VNF *** framework tackles the challenges in NFV performance *** framework components include chain graph generation,automatic deployment,and fine granularity *** design and implementation of each component have their *** the best of our knowledge,we make the first attempt to collect rules forming a knowledge base for generating reasonable chain *** deploys the generated chain graphs automatically,which frees the network operators from executing at least 391 lines of bash commands for a single *** diagnose the performance bottleneck,NfvInsight collects metrics from multiple layers of the software ***,we collect the network stack latency distribution ingeniously,introducing only less than 2.2%*** showcase the convenience and usability of NfvInsight in finding bottlenecks for both VNF chains and the underlying *** our framework,we find several design flaws of the network stack,which are unsuitable for packet forwarding inside one single server under the NFV *** optimization for these flaws gains at most 3x performance improvement.

关键词： network function virtualization(NFV) service chain performance bottleneck network stack latency

来源：评论

学校读者我要写书评

暂无评论

GenerOS: An asymmetric operating system kernel for multi-core systems

GenerOS: An asymmetric operating system kernel for multi-cor...

引用

24th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2010

作者： Yuan, Qingbo Zhao, Jianbo Chen, Mingyu Sun, Ninghui Key Laboratory of Computer System and Architecture Institute of Compute Technology Chinese Academy of Sciences Beijing China

ISBN: (纸本)9781424464432

Due to complex abstractions implemented over shared data structures protected by locks, conventional symmetric multithreaded operating system kernel such as Linux is hard to achieve high scalability on the emerging multi-core architectures, which integrate more and more cores on a single die. This paper presents GenerOS - a general asymmetric operating system kernel for multi-core systems. In principal, GenerOS partitions processing cores into application core, kernel core and interrupt core, each of which is dedicated to a specified function. In implementation, we conduct a delicate modification to Linux kernel and provide the same interface as Linux kernel so that GenerOS is compatible with legacy applications. The better performance of GenerOS mainly benefits from: (1) Applications run on their own cores with minimal interrupt and kernel support;(2) Every kernel service is encapsulated into a serial process so that there will be fewer contentions than conventional symmetric kernel;(3) A slim schedule policy is used in the kernel core to support schedule between system calls with low overhead. Experiments with two typical workloads on 16-core AMD machine show that GenerOS behaves better than original Linux kernel when there are more processing cores (19.6% for TPCH using oracle database management system and 42.8% for httperf using apache web server). © 2010 IEEE.

关键词： Linux

来源：评论

学校读者我要写书评

暂无评论

Indexing Techniques of Distributed Ordered Tables： A Survey and Analysis

引用

Journal of computer Science & technology 2018年第1期33卷 169-189页

作者： Chen Feng Chun-Dian Li Rui Li State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing 100190 China~TT ' University of Chinese Academy of Sciences Beijing 100049 China Tencent Inc. Beijing 100080 China

Many NoSQL （Not Only SQL） databases were proposed to store and query on a huge amount of data. Some of them like BigTable, PNUTS, and HBase, can be modeled as distributed ordered tables （DOTs）. Many additional indexing techniques have been presented to support queries on non-key columns for DOTs. However, there was no comprehensive analysis or comparison of these techniques, which brings troubles to users in selecting or proposing a proper indexing technique for a certain workload. This paper proposes a taxonomy based on six indexing issues to classify indexing techniques on DOTs and provides a comprehensive review of the state-of-the-art techniques. Based on the taxonomy, we propose a performance model named QSModel to estimate the query time and storage cost of these techniques and run experiments on a practical workload from Tencent to evaluate this model. The results show that the maximum error rates of the query time and storage cost are 24.2% and 9.8% respectively. Furthermore, we propose IndexComparator, an open source project that implements representative indexing techniques. Therefore, users can select the best-fit indexing technique based on both theoretical analysis and practical experiments.

关键词： database Not Only SQL (NoSQL) range query indexing

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：