Muutke küpsiste eelistusi

E-raamat: Annual Review of Scalable Computing [World Scientific e-raamat]

(Nus, S'pore)
  • Formaat: 228 pages
  • Sari: Series On Scalable Computing 3
  • Ilmumisaeg: 17-Apr-2001
  • Kirjastus: World Scientific Publishing Co Pte Ltd
  • ISBN-13: 9789812810182
Teised raamatud teemal:
  • World Scientific e-raamat
  • Hind: 137,68 €*
  • * hind, mis tagab piiramatu üheaegsete kasutajate arvuga ligipääsu piiramatuks ajaks
Annual Review of Scalable ComputingSuurem pilt
  • Formaat: 228 pages
  • Sari: Series On Scalable Computing 3
  • Ilmumisaeg: 17-Apr-2001
  • Kirjastus: World Scientific Publishing Co Pte Ltd
  • ISBN-13: 9789812810182
Teised raamatud teemal:
World Scientific e-raamatudRohkem infot World Scientific e-raamatute kohta
Raamatu kodulehekülg: http://www.worldscientific.com/worldscibooks/10.1142/4663#t=toc
The third volume in the Series on Scalable Computing, this book contains five new articles describing significant developments in the field. Included are such current topics as clusters, parallel tools, load balancing, mobile systems, and architecture independence.
Anatomy of a Resource Management System for HPC Clusters
1(31)
Introduction
1(3)
Computing Center Software (CCS)
2(1)
Target Platforms
3(1)
Scope and Organisation of this
Chapter
3(1)
CCS Architecture
4(13)
User Interface
4(1)
User Access Manager
5(1)
Scheduling and Partitioning
6(4)
Access and Job Control
10(2)
Performance Issues
12(1)
Fault Tolerance
12(1)
Modularity
13(4)
Resource and Service Description
17(7)
Graphical Representation
17(2)
Textual Representation
19(1)
Dynamic Attributes
20(2)
Internal Data Representation
22(1)
RSD Tools in CCS
23(1)
Site Management
24(2)
Center Resource Manager (CRM)
24(2)
Center Information Server (CIS)
26(1)
Related Work
26(1)
Summary
27(1)
Bibliography
28(4)
On-line OCM-Based Tool Support for Parallel Applications
32(31)
Introduction
33(1)
OMIS as Basis for Building Tool Environment
34(1)
Adapting the OCM to MPI
35(4)
Handling Applications in MPI vs. PVM
36(1)
Starting-up MPI Applications
36(2)
Flow of Information on Application
38(1)
Detection of Library Calls
39(1)
Integrating the Performance Analyzer PATOP with the OCM
39(6)
PATOP's Starter
39(1)
Prerequisites for Integration of PATOP with the OCM
40(1)
Gathering Performance Data with the OCM
40(2)
New Extension to the OCM - PAEXT
42(1)
Modifications to the ULIBS Library
43(1)
Costs and Benefits of using the Performance Analysis Tool
44(1)
Adaptation of PATOP to MPI
45(5)
Changes in the Environment Structure
45(1)
Extensions to ULIBS
45(1)
MPI-Specific Enhancements in PATOP
46(2)
Monitoring Overhead Test
48(2)
Interoperability within the OCM-Based Environment
50(5)
Interoperability
50(1)
Interoperability Support in the OCM
51(1)
Interoperability in the OCM-Based Tool Environment
51(1)
Possible Benefits of DETOP and PATOP Cooperation
52(1)
Direct Interactions
53(2)
A Case Study
55(3)
Concluding Remarks
58(1)
Bibliography
59(4)
Task Scheduling on NOWs using Lottery-Based Work Stealing
63(22)
Introduction
64(5)
The Cilk Programming Model and Work Stealing Scheduler
69(4)
Java Programming Language and the Cilk Programming Model
69(2)
Lottery Victim Selection Algorithm
71(2)
Architecture and Implementation of the Java Runtime System
73(3)
Architecture of the Java Runtime System
73(3)
Implementation of the Java Runtime System
76(1)
Performance Evaluation
76(3)
Applications
76(2)
Results and Discussion
78(1)
Conclusions
79(1)
Bibliography
80(5)
Transaction Management in a Mobile Data Access System
85(63)
Introduction
86(3)
Multidatabase Characteristics
89(10)
Taxonomy of Global Information Sharing Systems
90(1)
MDBS and Node Autonomy
91(1)
Issues in Multidatabase Systems
92(1)
MDAS Characteristics
93(4)
MDAS Issues
97(2)
Concurrency Control and Recovery
99(9)
Multidatabase Transaction Processing: Basic Definitions
102(1)
Global Serialisability in Multidatabases
103(1)
Multidatabase Atomicity / Recoverability
104(2)
Multidatabase Deadlock
106(1)
MDAS Concurrency Control Issues
107(1)
Solutions to Transaction Management in Multidatabases
108(17)
Global Serializability under Complete Local Autonomy
110(3)
Solutions using Weaker Notions of Consistency
113(2)
Solutions Compromising Local Autonomy
115(2)
Using Knowledge of Component Databases
117(1)
Global Serializability Based on Transaction Semantics
118(1)
Solutions under MDAS
118(4)
Solutions to Global Atomicity and Recoverability
122(3)
Application Based and Advanced Transaction Management
125(11)
Unconventional Transactions Types
126(1)
Advanced Transaction Models
126(2)
Replication
128(7)
Replication Solutions in MDAS
135(1)
Experiments with V-Locking Algorithm
136(4)
Simulation Studies
138(1)
System Parameters
139(1)
Simulation Results
139(1)
Conclusion
140(2)
Bibliography
142(6)
Architecture Inclusive Parallel Programming
148(64)
Introduction
148(9)
Architecture Independence - The Holy Grail
148(2)
Shared Memory Versus Distributed Systems
150(2)
Homogeneous Versus Heterogeneous Systems
152(3)
Architecture Independence Versus Inclusiveness
155(2)
Concurrency
157(7)
Threads and Processes
157(1)
Exclusion and Synchronization
158(2)
Atomicity
160(2)
Monitors and Semaphores
162(2)
Data Parallelism
164(7)
Vector Processors
164(2)
Hypercubes
166(1)
PRAM Algorithms
167(4)
Memory Consistency
171(6)
Shared Memory System
171(2)
Tuplespace
173(1)
Distributed Processing
174(2)
Distributed Shared Memory
176(1)
Tuple Locks
177(12)
The Need for Better Locks
177(4)
Tuple Locks
181(1)
Using Tuple Locks
182(3)
Bucket Location
185(1)
Homogeneous Systems
186(3)
Using Tuple Locks in Parallel Algorithms
189(9)
Gaussian Elimination
189(1)
Prime Numbers
190(2)
Fast Fourier Transform
192(3)
Heap Sort
195(3)
Tuples in Objects
198(5)
Objects and Buckets
198(1)
An Example
199(3)
Reflective Objects
202(1)
Towards Architecture Inclusive Parallel Programming
203(9)
Parallel Tasking
203(2)
Speculative Processing
205(1)
Efficient Implementation of Tuple Operations
206(2)
Tuple and Bucket Programming Styles
208(4)
Back Towards Architecture Independence
212
Püsilink: https://www.kriso.ee/db/9789812810182_pe.html