the Desirable Garfield

- 클리앙
- 디시인사이드
- 야놀자
-

The steady increases in processor speeds associated with Moore's Law has improved software performance for decades without necessitating significant changes in software designs or practices. Over the past several  years, however, the exponential growth in CPU speed has stalled. Increases in software performance now stem largely from exploiting parallel processing to exchange data reliably and scalably across high-speed interconnects, dynamically balance workload in computation grids, and efficiently synchronize access to shared resources. Researchers and practitioners rely on parallel processing to accelerate scientific discoveries and deliver value to users in a wide range of application domains, including high-performance scientific computing, weather forecasting, financial services, animation rendering, text mining, homeland security and enterprise content management.
 

무어의 법칙대로 프로세서 속도의 지속적인 증가는, 소프트웨어 설계나 관행에 특별한 변화를 필요로 하지 않고, 수 십년간 소프트웨어 성능을 개선시켜 왔습니다.
  

 Although parallel processors and interconnects continue to improve, it remains tedious and error-prone to develop complex application and infrastructure software that can meet challenging – and changing – user requirements. This situation has yielded a ‘parallel software crisis’, in which the hardware becomes ever more capable but the software remains hard to develop, debug, optimize and evolve. Much of the effort expended on parallel software is spent rediscovering core concepts such as coordination, communication, and synchronization, and reinventing common components such as active objects, dynamic load balancers, job schedulers, message brokers and notification engines. Moreover, despite advances in key technologies, such as concurrent programming languages, vectorizing and optimizing compilers, operation system clustering techniques and grid computing middleware, many software developers lack experience of how and when to best apply these technologies.
 

 Addressing the parallel software crisis therefore requires more than just adopting the latest technologies it requires learning and applying successful parallel software patterns that document recurring architectures and designs and convey proven parallel software structures, algorithms and best practices. Knowledge of patterns helps researchers and practitioners to avoid rediscovering and reinventing core concepts and common components of parallel software. Patterns can also explain how and when to best apply parallel technologies.


 Popular patterns such as Adapter, Bridge, Reactor and Strategy have captured and guided the designs of application and infrastructure software for two decades. Many of these patterns were identified initially by developers of object-oriented graphical user interface frameworks that work in contexts where quality factors like usability, extensibility and portability are paramount. In addition to these quality factors, developers of parallel software must also understand and apply patterns that work in context in which low latency and high throughput, reliability and scalability are paramount.

Over the year, isolated coverage of parallel software patterns has appeared in various conference proceedings and book. For example, those associated with the Pattern Languages of Programming ( PLoP ) conferences present patterns for scalable locking and communication. Likewise, the Pattern-Oriented Software Architecture( POSA) series of books presents patterns for pipeline parallelism, master/slave processing, distributed request brokering and dynamic resource management. Until Jorge Ortega-Arjona published this book on pattern for parallel software design, however, no single source provided such a broad and deep spectrum of architectural patterns, design patterns and common idioms for developing parallel software.

Popular patterns such as Adapter, Bridge, Reactor and Strategy have captured and guided the designs of application and infrastructure software for two decades. Many of these patterns were identified initially by developers of object-oriented graphical user interface frameworks that work in contexts where quality factors like usability, extensibility and portability are paramount. In addition to these quality factors, developers of parallel software must also understand and apply patterns that work in context in which low latency and high throughput, reliability and scalability are paramount.

 The patterns and idioms that Jorge present in this book help to resolve key parallel software challenges such as coordinating interaction between concurrently executing task, partitioning parallel algorithms and data to improve performance substantially, and minimizing synchronization overhead in local and distributed shared memory. In addition to describing the structure and functionality of essential parallel software patterns and idioms, Jorge also presents many examples from a range of applications domains, including high-performance scientific computing, image processing and animation rendering. Moreover, Jorge’s detailed case studies extend the book beyond a catalog of parallel software patterns to provide keen insights into parallel software design processes and methods that help alleviate key accidental and inherent complexities in parallel software development projects.

For parallel software development to develop from an art to an engineering discipline, successful practices and design expertise must be documented systematically and disseminated broadly. My colleagues and I have documented and applied patterns in a wide range of distributed and parallel application and infrastructure software, including the ACE, TAO and Zircomp middleware. We’ve found that studying and applying patterns helps to:

n  Facilitate reuse of architecture and design artifacts, which reduces the effort required to develop high-quality parallel software frameworks and application components. These patterns can be reused even when reuse of algorithms, implementations, interfaces or detailed design is not feasible due to heterogeneous software and hardware platforms.

n  Document ‘best practices’ of parallel software systems, which have traditionally resided in the minds of expert developers or buried within complex source code. Capturing the most useful strategies and tactics of parallel software in terms of patterns makes the learning curve for new developers more gentle, by giving them good role models for developing parallel software applications and infrastructure.

n  Preserve important design information, which is often lost over time in conventional development processes, causing increased maintenance costs and software defects. Software evolution effort can thus be reduced significantly by documenting the intent, structure and behavior of parallel software components in terms of the patterns they reify, as well as explaining how and when to best apply these components in various contexts.

n  Guide design choices for new systems, since patterns capture proven experience in a from that can be used to address new design challenges. By understanding the potential traps and pitfalls in their domains, developers can select suitable parallel software architectures, protocols and platform features without wasting time and effort implementing solutions that are known to be inefficient or error-prone.

 

 A thorough understanding of the parallel software patterns, processes and methods in Jorge’s book will help you develop better parallel software applications and infrastructure likewise. If you want a thorough coverage of the key pattern-oriented software architectures that are shaping the next-generation of parallel software, then read this book. I’ve learned much from it and I’m confident that you will too.

  

Douglas C. Schmidt

Nashville, Tennessee, USA

 

 

 

- 불펌 하시면 안됩니다. ^^ [ yougom.com ] -


서문

저자서문
저자에 대해
감사글
기고자 일대기

챕터 1.
소프트웨어 패턴
소프트웨어 패턴의 개념
패턴 서술, 구성, 분류
요약

챕터 2.
병렬 프로그래밍에 쉽게 접근하기
병렬 프로그래밍
병렬 프로그램 수행에 영향을 미치는 요인
병렬 프로그래밍의 장점과 단점
요약

챕터 3.
병렬 프로그래밍을 위한 아키텍쳐 패턴
병렬 파이프 & 필터
병렬 레이어
선형 요소 전달하기
매니저-워커
공용 자원
요약

챕터 4.
통신 컴포넌트의 디자인 패턴
공용 가변 파이프
복합적인 지역( Local ) 호출
메시지 전달 파이프
복합적인 원격 호출
공용 가변 채널
메시지 전달 채널
지역 랑데뷰
원격 랑데뷰
요약

챕터 5.
동기화 메커니즘의 표현식 ( Idiom )
세마포어
크리티컬 영역
모니터
메시지 전달
원격 프로시저 호출
요약

챕터 6.
두가지 사례 연구
혈관 세분화
조절 가능한 3D 그리드 기반 오일러( 기체역학의 ) 프로그램
요약

챕터 7.
병렬 소프트웨어 디자인
일반적인 병렬 소프트웨어 디자인 프로세스
패턴 기반 병렬 소프트웨어 디자인 방법
문제 분석
조직화 ( Coordination ) 설계 - 아키텍쳐 패턴
통신 ( Communication ) 설계 - 디자인 패턴
세부 설계 - 표현식 ( Idiom )
구현과 진화
요약

챕터 8.
병렬 소프트웨어 아키텍쳐
병렬 소프트웨어 아키텍쳐의 정의
병렬 소프트웨어 디자인
요약

챕터 9.
병렬 프로그래밍 패턴의 방향
소프트웨어 설계의 상황
설계 지식과 기술
병렬 소프트웨어 시스템의 실제 서술
병렬 소프트웨어 설계 측정의 필요성

용어 사전

표기법

참조 문헌

패턴 색인

색인


















------------------------------------ 원문 -----------------------------------------

- 무단 불펌 하시면 안됩니다. ^^ [ yougom.com ] -


[ 앞표지 ]

Patterns for Parallel Software Design
병렬 소프트웨어 설계를 위한 패턴

Jorge Luis Ortega-Arjona 
저자. 조지 루이스 오르티가-아르조나 


[ 뒷표지 ]

Patterns for Parallel Software Design is essential reading for developers looking to understand patterns for parallel programming.
병렬 소프트웨어 설계 패턴은 병렬 프로그래밍 패턴들을 습득하려는 개발자들이 필수적으로 읽어야 합니다.



Written from an architectural point of view, Patterns for parallel Software Design presents a pattern-oriented software architecture approach to parallel software design providing solutions in concurrent and distributed programming, based on existing design knowledge.

저자는 아키텍쳐 관점에서, 병렬-소프트웨어-설계-패턴을 기존 설계 지식에 기반하는 동시처리와 분산처리 프로그래밍의 해결책으로 병렬-소프트웨어-설계를 패턴-지향-소프트웨어-아키텍쳐로 접근 합니다.



A pattern-oriented approach to parallel software design is not a design method in the classic sense,
but a new way of managing and utilizing existing design knowledge for designing parallel programs.
 Using this approach leads to parallel software systems that are modular, adaptable, understandable and evolvable.
Thus, this method aims to enhance not only build-time properties of parallel systems, but also their run-time properties.

패턴 지향적 병렬 소프트웨어 설계는 클래식한 설계법이 아니지만, 병렬 프로그래밍을 설계하는 기존 설계 지식을 활용하고 관리하는 새로운 방법입니다. 이를 활용해 접근한다면 모듈화되고 적응력( adaptable )있으며 이해하고 쉽고 진화적인 병렬 프로그램 설계로 이끌어 낼 수 있습니다. 이와 같은 방법들은 병렬 시스템의 빌드타임 속성뿐만 아니라, 런타임 속성 향상에 도움을 줍니다.



Key topics include :
 - Use of known solutions in concurrent and distributed programming, applied to the development of parallel programs.
 - Significant architectural patterns that describe how to divide an algorithm and/or data to find a suitable partition, and hence, link it with a programming structure that allows for such a division.
 - Delivers proven solutions to the problems faced by parallel programmers.

주요 주제는 다음과 같습니다.
 - 동시 및 분산 프로그래밍에 알려진 해결책은 병렬 프로그램 개발에 사용됩니다.
 - 어떻게 알고리즘과 데이터를 구분할 것인가, 어떻게 적절한 분할 영역을 찾을 것인가에 대해 묘사해주는 중요한 아키텍저적인 패턴 입니다. 게다가, 분할 구조로 하는 프로그래밍과 관련이 있습니다.
 - 병렬 프로그래밍에 직면한 문제들의 입증된 해결책을 제시합니다.



Coverage is aimed at developers new to parallel programming, who require a base to understand parallel software design and implementation for future parallel platforms. Patterns for Parallel Software Design is an essential must-have guide for developers and programmers who want to solve unique design problems.

이 책은 병렬 소프트웨어 설계의 기본적인 이해와 미래적 병렬 플랫폼을 위한 구현을 요구하는 개발자에게 새로운 병렬 프로그래밍을 하도록 하는 것입니다. 병렬 소프트웨어 설계 패턴은 고유의( Unique ) 설계 문제들을 풀고 싶어 하는 개발자들과 프로그래머들에게 필수로 지니고 있어야 할 안내서입니다.