DOC《课程编号：2122003》

课程编号:2122003 课程名称:高级计算机体系结构 英文名称:Advanced Computer Architecture 总学时:

32 学分:2 预修课程:计算机体系结构 内容简介:本课程介绍计算机系统的设计思想、设计原理、设计技术的一门核心课程.

是计算机系硕士研究生和博士生的一门学位课程,也是信息学院、电气学院相关专业研究生的重要选修课程.主要介绍现代计算机处理器中流水线技术、指令级并行性开发技术、多处理器计算机设计技术.具体内容包括: Ch1. 计算机系统结构引论 Ch2. 基本流水线技术 Ch3. 指令级并行性及动态开发技术 Ch4. 指令级并行性的局限性 Ch5. 多处理器器和线程级并行性 Ch6. 向量处理器 Ch7. VLIW和EPIC的硬件和软件 英文简介:This course is opened for MS students and Ph.D students in computer science and engineering the advanced technologies in computer architecture. It will mainly focus on the technologies in computer design of pipelining, Instruction-Level Parallelism exploitation and multi-processors. This course will cover the following topics: Ch1. Fundamentals of computer design Ch2. Pipelining: basic and intermediate concepts Ch3. Instruction-Level Parallelism and its exploitation. Ch4. Limits on Instruction-Level Parallelism Ch5. Multiprocessors and Thread-Level Parallelism Ch6. Vector Processors Ch7. Hardware and Software for VLIW and EPIC 教材或讲义:John L. Hennessy, David A. Patterson,《Computer Architecture―A Quantitative Approach》,4th Edition, China Machine Press,

2007 参考书:1. 郑纬民,汤志忠,汪东升,《计算机体系结构――量化研究方法》(第3版)中文版.电子工业出版社,2004年. 撰写人:姜晓红 课程编号:2122018 课程名称:高级计算机网络 英文名称:Advanced Computer Network 总学时:

32 学分:2 预修课程:计算机网络(本科) 内容简介:全面介绍了快速变化的有线和无线计算机互连网络世界,提供尽量全面的内容,强调网络技术的原理和概念.在网络技术不断进步,网络的实现变得越来越复杂和专业化的情况下,这些基本原理和概念始终万变不离其宗,或者说变化很少,突出介绍网络领域中的一些重要标准,介绍网络技术中的有关理论,并通过真实例子来加以说明.从基本的概念入手,帮助学生熟悉已经学过的计算机网络的相关内容,并随后进行深入的网络技术的内容,这些内容包括了:数据编码与传输,数据链路协议,局域网技术,家庭网络技术,广域网技术,网络互连技术等.同时,根据计算机网络的发展,尤其是因特网的发展,介绍了因特网技术的发展与变化.课程以计算机网络原理为基础,以应用为导向,以发展与变化为指导,阐述当今高级计算机网络的各种理论、思想和方法,力求使学生了解计算机网络技术的发展,为从事高水平研究打好基础. 英文简介:Advanced Network is organized into different topics for ease of teaching and learning. It first presents introductory material that lays the foundation and provides a framework for the more detailed material in subsequent chapters. This will be helpful to the students for their advanced study. Then it delves into the fundamental technology that underlies all telecommunications and networking, and provides the technical foundation for networks including how data are coded for transmission on a network;

analog and digital signals, transmission methods, and the way they are combined in various network systems, the protocols that are used on data links and so on. Further it examines networks in detail. LANs are studied first, followed by WANs: various topologies, access control techniques, protocols, and routing techniques are explained for LANs;

the intricacies and limitations of the technologies of real-world networks are covered;

packet data networks and the way traffic is routed in a WAN are explained. Then it deals with the ways in which networks of all types are interconnected into internets. And it relates various Internet applications and specifically, those of the WWW and the technologies that underlie its operation are explored. Last it examines network security, network design, and network management. Various types of management reporting of network operations are shown. 教材或讲义:自编讲义 参考书:1. ANDREW S.TANENBAUM,VRIJE UNIVERSITEIT,AMSTERDAM,THE NETHERLANDS著,潘爱民译,计算机网络(第4版)(中文版),清华大学出版社,2004年8月 撰写人:陈天洲 课程编号:

2124059 课程名称: 多核计算 英文名称: Multi-core computing 总学时:

32 学分:

2 预修课程:高级计算机体系结构(硕士)、高级操作系统(硕士) 内容简介:随着计算机技术的日新月异的发展,原有的芯片设计已经基本达到了其性能的瓶颈,开发新的芯片结构成为了推动计算机技术继续发展的立足点.而多核芯片设计就是现今SOC设计的一个最重要的方向.多核体系打破了原有的单芯片单处理核心的设计,在同一芯片中融入了多个处理核心.这种设计大大增强了处理器的处理能力,真正实现了多线程设计.同时,这种设计又不同于多处理器的体系设计.多个处理核心之间的通信的代价较多处理器设计要小的多,整体系统的复杂度因而也可降低许多,并带来了成本的降低以及系统稳定性的提高.本课程是在研究生的体系结构课程和超大规模集成电路基础上做进一步提高,结合操作系统、编译原理、编程方法等交叉学科,介绍在体系结构上的多核理论与相关学科对此的相应发展.课程以多核体系结构、多核芯片现状、多核系统软件、多核编译优化、多核编程方式等为主线,阐述当今多核体系的基本理论、思想和方法,力求使研究生们了解世界前沿的体系结构技术,为从事高水平研究打好基础. 英文简介:With the development of new computing technology, original computer architecture design has already reached its bottle-neck, and developing new chip architecture comes to be the stand of future development of the computer technology. Multi-Core design is one of the most important direct in the System on Chip domain. This architecture breaks our original design with two or more cores in one chip. This can rapidly improve the computing ability of CPU and achieve real multi-thread design. Furthermore, it is quite different from the SMP architecture. Its communication cost between cores will be much less than the SMP, which brings down the whole system'

s complexity、the cost of HW and increases the stability of the whole system. This course is an inter-discipline of architecture area, operating system (OS), compiling principle and programming style. It will introduce the fundamentals of Multi-Core and relative domains'

development on it. The basic principle、idea and methods of Multi-core architecture design will be illustrated through several categories, i.e.,Multi-Core architecture , actuality of Multi-Core chip, Multi-Core system software, Multi-Core compiling optimize, Multi-Core programming style. Graduate students can master the latest theory and methods of modern image processing from this course for incoming research work. 教材或讲义:自编讲义 参考书:Shameem Akhter and Jason Roberts,《Multicore programming》,intel press,2006年4月 撰写人:陈天洲 课程编号:

2124012 课程名称: 网格计算 英文名称: Grid Computing &

Distributed System 总学时:

32 学分:2 预修课程:无 内容简介:本课程主要围绕网格计算介绍新一代的分布式计算与互联网技术,具体内容包括计算网格(Computational Grid)、Web 服务(Web S........