Dark2 reading instructions for 3rd edition of the book
Sections marked RT are "read through" sections, i.e., understand the overall message but do not concentrate on any details. All the slides shown in class are to be considered as course material, unless stated otherwise.
Caches
Chapter 5 up until pg432
(There are lots of formulas of a fairly simple nature. I expect you to be able to derrive them from your understanding of the text, but not to know them by heart)
Memory and Optimizations
The rest of Chapter 5
Important sections: Section 5.4, 5.5 and 5.10
Fig 5.26 is RT
From 5.13 is RT
Instruction set
1.6
1.11 RT
2.2 2.5 2.12 2.14 and 2.15
All other sections of Ch2: RT
Pipelining
Appendix A : A.1 - A.5
A.6: RT
Advanced pipelining
3.1 (partly a repeat of appendix A)
3.2 - 3.3 (understand the essence of Tomasulo, but not implementation details)
3.4 - 3.6 (Multiple Inst Issue with Dynamic Scheduling is RT)
3.7 - 3.8: RT
Software approaches to ILP
4.1 - 4.4 (skip p320-329, paragraphs about trace scheduing and superblocks are RT)
4.5 - 4.7
Embedded processors
The generic information on the slides are included (i.e., not details about specific CPUs etc.)
1.8
2.13 RT
5.14 RT
Slides are included (not details about specific CPUs)
Network processors
The generic information from the slides are included (i.e., not implementation details etc.)
Introduction to MPs
6.1 RT
6.3
6.5
6.8
Implementing MPs
1.3
1.7
6.9
Large MPs
5.15 RT
6.10-6.11
The WildFire paper is RT
Programming MPs
6.4
Synchronization
6.7
The Scott paper (except the issues that deal with timeout)
Future
The Agarwal paper (the trends but not the formulas)