检索结果-内蒙古大学图书馆

7th International Green and Sustainable Computing Conference, IGSC 2016

作者： Hu, Jiuchuan Fan, Dongrui Li, Danping Yan, Long Ye, Xiaochun School of Computer and Information Technology of Beijing Jiaotong University Beijing100044 China State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China Institutes of Information Engineering Chinese Academy of Science Beijing100093 China

ISBN: (纸本)9781509051175

To reduce data fetching latency in multi-core processor, a new on-chip cache model, Percolation Cache Hierarchy(PCH) that is a hybrid combination of convention on-chip caches, and a new data migration schema, called Percolation Data Migration Schema(PDMS), for moving data massively on PCH to enrich hit possibility and reduce data moving latency, are introduced in this paper. Basing on the percolation data migration schema, a new on-chip data environment, Just In-Time Data Environment(JITDE), can be setup on the PCH to exploit the properties of data locality to increase hit rate. © 2016 IEEE.

关键词： Solvents

来源：评论

学校读者我要写书评

暂无评论

Efficient runtime support for embedded MPSoCs

Efficient runtime support for embedded MPSoCs

引用

International Conference on Embedded computer Systems: architectures, Modeling and Simulation (IC-SAMOS)

作者： Dimitris Theodoropoulos Polyvios Pratikakis Dionisios Pnevmatikatos Computer Architecture and VLSI Systems Laboratory Institute of Computer Science Foundation for Research and Technology-Hellas Crete Greece

ISBN: (纸本)9781479901043

Recently, many software runtime systems have been proposed that allow developers to efficiently map applications to contemporary consumer electronic devices and high-performance academic processing platforms. Most of these runtime systems employ advanced scheduling techniques for automatic task assignment to all available processing elements. However, they focus on a particular environment and architecture, and it is not easy to port them to reconfigurable embedded MPSoCs. As a consequence, in the embedded community, researchers implement hardwired application-specific task schedulers, which can not be used by other embedded MPSoCs. To address this problem, in this paper we propose a lightweight runtime software framework for reconfigurable shared-memory MPSoCs, that integrate a master embedded processor connected to slave cores. Similarly to many of the aforementioned advanced runtime systems, we adopt a task-based programming model that uses simple, pragma-based annotations of the application software, in order to dynamically resolve task dependencies. Our runtime system supports heterogeneity in the hardware resources, and is also low-overhead to account for possible limitations in their processing capabilities and available on-chip memory. To evaluate our proposal, we have prototyped an MPSoC with seven slaves to a Xilinx ML605 FPGA board. We run three micro-benchmarks that achieve a performance speedup of 3.8x, 7x and 5.8x, and energy consumption of 27%, 14% and 18% respectively, compared to a single-core baseline system with no runtime support.

关键词： Runtime Software Hardware Real-time systems Field programmable gate arrays Processor scheduling Programming

来源：评论

学校读者我要写书评

暂无评论

Reward Modeling Requires Automatic Adjustment Based on Data Quality

Reward Modeling Requires Automatic Adjustment Based on Data ...

引用

2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024

作者： Wang, Binghai Zheng, Rui Chen, Lu Xi, Zhiheng Shen, Wei Zhou, Yuhao Yan, Dong Gui, Tao Zhang, Qi Huang, Xuanjing School of Computer Science Fudan University China Baichuan Inc. Beijing China Institute of Modern Languages and Linguistics Fudan University China Key Laboratory of Intelligent Information Processing Fudan University Shanghai China

ISBN: (纸本)9798891761681

In Reinforcement Learning from Human Feedback (RLHF), the reward model plays a crucial role in aligning language model outputs with human values. The human preference data used to train the reward model consists of a prompt and a response pair, with humans annotating which response better aligns with human value preferences. Due to the complexity and subjectivity of the annotation task, multiple organizations including OpenAI and Anthropic report significant noise in the human preference datasets, leading to instability and deviation in reward model training from human values. We discover that the difference in scores assigned to response pairs by the reward model effectively indicates the quality of data, and data of varying qualities show significant distinctions in reward model training. We introduce a method that automatically adjusts reward modeling based on data quality, reducing the impact of noise and making full use of dataset. Experiments on multiple human preference datasets demonstrate that our method stabilizes reward model training and significantly enhances the alignment performance of RLHF. © 2024 Association for Computational Linguistics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Hunting Trojan Horses

Hunting Trojan Horses

引用

ASID'06: 1st Workshop on Architectural and System Support for Improving Software Dependability

作者： Moffie, Micha Cheng, Winnie Kaeli, David Zhao, Qin Computer Architecture Research Laboratory Northeastern University Boston MA 02115 United States Computer Science and Artificial Intelligence Lab. Massachusetts Institute of Technology Cambridge MA 02139 United States Singapore-MIT Alliance National University of Singapore Singapore 117576 Singapore

ISBN: (纸本)1595935762

HTH (Hunting Trojan Horses) is a security framework developed for detecting difficult types of intrusions. HTH is intended as a complement to anti-virus software in that it targets unknown and zero-day Trojan Horses and Backdoors. In order to accurately identify these types of attacks HTH utilizes runtime information available during execution. The information collected includes fine-grained information flow, program execution flow and resources *** this paper we present Harrier, an Application Security Monitor at the heart of our HTH framework. Harrier is an efficient run-time monitor that dynamically collects execution-related data. Harrier is capable of collecting information across different abstraction levels including architectural, system and library APIs. To date, Harrier is 3-4 times faster than comparable information flow tracking *** the collected information, Harrier allows for accurate identification of abnormal program behavior. Preliminary results show a good detection rate with a low rate of false positives. Copyright 2006 ACM.

关键词： computer software

来源：评论

学校读者我要写书评

暂无评论

Photo rendering with swarms: From figurative to abstract pherogenic imaging

Photo rendering with swarms: From figurative to abstract phe...

引用

IEEE Workshop on Affective Computational Intelligence (WACI)

作者： Carlos M. Fernandes Antonio Mora Juan Julián Merelo Carlos Cotta Agostinho C. Rosa Department of Computer Architecture and Technology Systems and Robotics Institute University of Granada Spain Department of Computer Architecture and Technology University of Granada Spain Department of Computer Languages and Computer Science University of Malaga Spain Systems and Robotics Institute Technical University of Lisbon Portugal

This paper describes the use of ant colony models for photographic rendering, that is, for the depiction or interpretation of photographic images. A description of a previously proposed ant-based edge detection method called pherographia is given in order to contextualize this paper's proposal. Then, the application of the KANTS algorithm to photo-rendering is introduced. KANTS is ant-based clustering algorithm described by a set of equations that model the local behavior of simple units (ants). These units represent data samples. When moving on a heterogeneous 2-dimensional lattice of vectors, global behavior emerges from the local interaction between the ants, which self-organize into clusters. In this paper, a data set consists of the RGB values of a digital (or digitalized) photograph. Two different approaches for the initialization of the environment are tested, each one resulting in radically different output images: when the environment is initialized with the RGB values of the same image that is used to generate the data set, the final result is very similar to the original photo; a random initialization of the environment leads to an abstract interpretation of the photo. The results are contextualized within previous swarm artworks by the authors.

关键词： Vectors Art Mathematical model Equations Lattices Image color analysis Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Formula-Oriented Compositional Minimization in Model Checking

Formula-Oriented Compositional Minimization in Model Checkin...

引用

2010 19th IEEE Asian Test Symposium(第19届IEEE亚洲测试技术学术会议 ATS 2010)

作者： Bowen Chen Haihua Shen Wenhui Zhang Key Laboratory of Computer System and Architecture Institute of Computing Technology Chinese Acade Key Laboratory of Computer System and Architecture Institute of Computing Technology Chinese Acade State Key Laboratory of Computer Science Institute of Software Chinese Academy of Sciences Beijin

This paper presents a new approach to reduce finite state machines with respect to a CTL formula to alleviate state explosion problem. Reduction is achieved by removing parts useless to the formula of original machines. The main contribution of this paper is to exploit relations among subformulas of the CTL formula so as to gain more reduction, as well as to extend traditional pruning method, which handles only existential formulas, to handle universal formulas. Based on this kind of reduction, verification of a large system, which usually consists of several components, can be done by evaluating properties on a reduced version of the system, which is built by composing components of the system one by one while doing reduction after each composition. Experimental results show the effectiveness of the approach. Especially when a property is written in a more detailed way, that is to describe the system part by part, the approach has a great potential.

关键词： Formula-Oriented Compositional Minimization CTL Model Checking Formal Verification

来源：评论

学校读者我要写书评

暂无评论

ZnG: Architecting GPU Multi-Processors with New Flash for Scalable Data Analysis

ZnG: Architecting GPU Multi-Processors with New Flash for Sc...

引用

Annual International Symposium on computer architecture, ISCA

作者： Jie Zhang Myoungsoo Jung Computer Architecture and Memory Systems Laboratory Korea Advanced Institute of Science and Technology (KAIST)

ISBN: (数字)9781728146614

ISBN: (纸本)9781728146621

We propose ZnG, a new GPU-SSD integrated architecture, which can maximize the memory capacity in a GPU and address performance penalties imposed by an SSD. Specifically, ZnG replaces all GPU internal DRAMs with an ultra-low-latency SSD to maximize the GPU memory capacity. ZnG further removes performance bottleneck of the SSD by replacing its flash channels with a high-throughput flash network and integrating SSD firmware in the GPU's MMU to reap the benefits of hardware accelerations. Although flash arrays within the SSD can deliver high accumulated bandwidth, only a small fraction of such bandwidth can be utilized by GPU's memory requests due to mismatches of their access granularity. To address this, ZnG employs a large L2 cache and flash registers to buffer the memory requests. Our evaluation results indicate that ZnG can achieve 7.5× higher performance than prior work.

关键词：

来源：评论

学校读者我要写书评

暂无评论

GraphTensor: Comprehensive GNN-Acceleration Framework for Efficient Parallel Processing of Massive Datasets

GraphTensor: Comprehensive GNN-Acceleration Framework for Ef...

引用

International Symposium on Parallel and Distributed Processing (IPDPS)

作者： Junhyeok Jang Miryeong Kwon Donghyun Gouk Hanyeoreum Bae Myoungsoo Jung Computer Architecture and Memory Systems Laboratory Korea Advanced Institute of Science and Technology (KAIST)

We present GraphTensor, a comprehensive open-source framework that supports efficient parallel neural network processing on large graphs. GraphTensor offers a set of easy-to-use programming primitives that appreciate both graph and neural network execution behaviors from the beginning (graph sampling) to the end (dense data processing). Our framework runs diverse graph neural network (GNN) models in a destination-centric, feature-wise manner, which can significantly shorten training execution times in a GPU. In addition, GraphTensor rearranges multiple GNN kernels based on their system hyperparameters in a self-governing manner, thereby reducing the processing dimensionality and the latencies further. From the end-to-end execution viewpoint, GraphTensor significantly shortens the service-level GNN latency by applying pipeline parallelism for efficient graph dataset preprocessing. Our evaluation shows that GraphTensor exhibits 1.4× better training performance than emerging GNN frameworks under the execution of large-scale, real-world graph workloads. For the end-to-end services, GraphTensor reduces training latencies of an advanced version of the GNN frameworks (optimized for multi-threaded graph sampling) by 2.4×, on average.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Avalon: Building an operating system for robotcenter

arXiv

引用

arXiv 2018年

作者： Xu, Yuan Yan, Zhiyuan Wang, Sa Yang, Cheng Xiao, Qingsai Bao, Yungang State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Science

This paper envisions a scenario that hundreds of heterogeneous robots form a robotcenter which can be shared by multiple users and used like a single powerful robot to perform complex tasks. However, current multi-robot systems are either unable to manage heterogeneous robots or unable to support multiple concurrent users. Inspired by the design of modern datacenter OSes, we propose Avalon, a robot operating system with two-level scheduling scheme which is widely adopted in datacenters for Internet services and cloud computing. Specifically, Avalon integrates three important features together: \(1\) Instead of allocating a whole robot, Avalon classifies fine-grained robot resources into three categories to distinguish which fine-grained resources can be shared by multi-robot frameworks simultaneously. \(2\) Avalon adopts a location based resource allocation policy to substantially reduce scheduling overhead. \(3\) Avalon enables robots to offload computation intensive tasks to the *** have implemented and evaluated Avalon on robots on both simulated environments and real world. Copyright © 2018, The Authors. All rights reserved.

关键词： Scheduling

来源：评论

学校读者我要写书评

暂无评论

Evolution from shark to spark SQL: Preliminary analysis and qualitative evaluation 6th

Evolution from shark to spark SQL: Preliminary analysis and ...

引用

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

作者： Tian, Xinhui Lu, Gang Zhou, Xiexuan Li, Jingwei State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China Beijing Academy of Frontier Science and Technology Beijing China

ISBN: (纸本)9783319290058

Spark is a general distributed framework with the abstraction called resilient distributed datasets (RDD). Database analysis is one of the main kinds of workloads supported on Spark. The SQL component on Spark has evolved from Shark to Spark SQL, while the core components of Spark also have evolved a lot comparing with the original version. We analyzed on which aspects Spark have made efforts to support many workloads efficiently and whether the changes make the support for SQL achieve better performance. © Springer International Publishing Switzerland 2016.

关键词： computers

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：