Class
|
Topic
|
Reading
Some Appendices are located on the publisher's website
|
Milestones
|
Dynamic Branch Prediction 4/2, 4/4
|
What this class is about
|
Review of pipelining
|
Section 3.1. If the review in lecture isn't jogging your memory, consult Appendix C.1-C.4.
|
Branch prediction
|
Appendix C, pp. 26-30; Sections 3.3 and 3.9, pp. 202-208
Project report guidelines,
and sample project report
|
Memory Hierarchy 4/9
|
Advanced caching techniques
|
Section 2.2
|
Execution Cores 4/11, 4/16, 4/18
|
Superscalars and static scheduling
|
|
Overview of dynamic scheduling
|
Section 3.4 to p.170, Section 2.8
|
Tomasulo's algorithm
|
Section 3.4, p.170 to its end;
Section 3.5
|
R10000-style dynamic scheduling (a physical register pool)
|
The
Smith/Sohi article for superscalars in a nutshell.
In the R10000 article
read from register mapping,
p. 32, through Register files, p. 35.
|
|
Prepare for the first midterm 4/23
|
|
|
Discuss the midterm and cover any outstanding material.
|
|
|
Midterm 1 4/25
|
Exam is 1.5 hours
|
|
Branch prediction homework due Friday, April 26, by midnight
|
Multiprocessors 4/30, 5/2, 5/7, 5/9
|
Overview of multiprocessing
|
Section 5.1
|
Cache coherency
|
Section 5.2, Section 5.4
|
Synchronization
|
Section 5.5
|
|
Wrap-up multiprocessing
|
|
Consistency models Brandon Myers, our TA
|
Consistency Primer, pgs. 2-4, 17-27, 37-47
|
|
Multithreading 5/14
|
An Overview of Multithreading and Simultaneous Multithreading
|
Section 3.5 and the SMT paper
|
First coherency milestone due, Thursday, May 14, before class time.
|
MP Scaling 5/16
|
MP Scaling for graph algorithms
Jacob Nelson (CSE grad student)
|
This paper,
"Cora: A Latency-Tolerant Runtime for Large-Scale Irregular Applications" is just FYI.
You're not responsible for reading it. This paper is currently under submission, so please don't distribute outside class.
|
|
Data Parallelism 5/21
|
GPUs, Derek Hower, AMD
|
Section 4.1, 4.2, 4.4
|
Combining ILP and TP 5/23, 5/28
|
Dataflow Machines
|
After reading them over, I don't think any of the papers on the
early dataflow machines are appropriate for classroom use. There
are no general overview papers. So just listen to the lecture.
|
|
Wavescalar architecture and implementation
|
Skim The WaveScalar Architecture and
An overview of the WaveScalar implementation.
Use this to reinforce what we discuss in lecture; don't pay much attention to any
new material covered in these papers.
|
|
|
FPGAs 5/30
|
FPGAs -- Silicon for Procrastinators, Andrew Putnam (Microsoft Research, formerly CSE Grad Student)
|
This paper,
"Performance and Power of Cache-Based Reconfigurable Computing", is just FYI.
You're not responsible for reading it.
|
|
The Wave of the Future 6/4
|
Dark Silicon and the End of Multicore Scaling, Hadi Esmaeilzadeh (CSE Grad Student)
|
This paper,
"Dark Silicon and the End of Multicore Scaling", is just FYI.
You're not responsible for reading it.
|
Architectures for Approximate Computing, Adrian Sampson (CSE Grad Student)
|
This paper,
Architecture Support for Disciplined Approximate Computing, is just FYI.
You're not responsible for reading it.
|
Coherency homework due, Tuesday, June 4, before class.
|
|
|
Midterm 2 6/6 10:30
|
Exam is 2 hours
|
|