COMPUTER ARCHITECTURE December
3, 2002
COMPREHENSIVE
EXAM INFORMATION
Day
& Time and Location |
Thursday
December 19th, 2002 @ 1:10pm-4pm in 717 Hamilton Hall |
Administered
By |
Professor
Tony Jebara, jebara@cs.columbia.edu |
Office
Hours |
Dec.
5 4:30-6pm Dec. 16 3:00-4:30pm, Dec. 17 1:30-3pm or by appointment |
TAs |
Description: The exam will focus on the MIPS Architecture
but explore the various concepts in CompOrg and CompArch including: digital
logic, implementation of arithmetic logic unit, binary numbers, representation
of negative numbers in a computer, floating point numbers, basic machine
instructors for a MIPS RISC-type computer, assembly language programming,
implementations of basic computer under various clocking assumptions,
pipelining. Memory hierarchy: caches and virtual memory. In addition,
Advanced Pipelining and Instruction Level Parallelism will be covered.
Texts: Topics from the following books
will be covered. Note the specific chapters.
David A.
Patterson and John L. Hennessy, Computer Organization and Design: the
Hardware/Software Interface, Morgan Kaufmann, San Mateo, CA (1998). (2nd
Edition):
CHAPTERS 1,2,3,4,5,6,7 and APPENDICES A,B.
John L. Hennessy and David A. Patterson, Computer
Architecture: A Quantitative Approach, Morgan Kaufmann, San Mateo, CA
(1996). (2nd Edition)
BEYOND PATTERSON & HENNESSY
CONCEPTS SO YOU ONLY
REALLY NEED TO READ CHAPTER
4.1-4.4 FROM H&P IN ADDITION…
Topics
Covered in P&H:
•Ch.1: RISC/CISC, Stored-Program, 5-Parts,
Hierarchy, bits
•Ch.2: Throughput, CPU Time, CPI, Arithmetic
Mean, Geometric Mean, SPEC,
Relative/Peak MIPS
•Ch.3: Operands, Registers, MIPS Programming,
Assembly, R,I,J-Types,
Pseudo-Instructions,
Loops, Pointers, Stack, Functions, Caller-Callee
Conventions, ASCII
Constant, Immediate, Addressing Styles, Link/Load
•Ch.4: Binary, 2’s Complement, Floating Point
#’s Logic, Addition,
Subtraction,
Overflow, Ripple-Carry ALU, Multiplication, Division
•Ch.5:
Registers, Memory, Clocking, Single-Cycle Datapath, Control,
lw,sw,beq,R-type,j,
Performance, Multi-Cycle Datapath, Control,
FSMs, MicroProgramming, Exceptions,
•Ch. 6: Pipelining, Hazards, Reordering Code,
Pipelined Datapath, Control,
Data Hazards,
Forwarding, Stalling, Branch Hazards, Prediction,
Scheduling,
Exceptions
•Ch.7: Memory Hierarchy, Direct-Mapped Caches,
Hits, Misses, MissRate,
Performance, Memory
Design, Associative Caches, Multi-Level Caches,
Virtual Memory,
Pages, Translation-Lookaside Buffer
•Appendices: SPIM, Digital Logic, Sequential, Clocks
Extra
Topics Covered in H&P:
•Ch.4.1: Instruction Level Parallelism
•Ch.4.2: Overcoming Data Hazards with Dynamic
Scheduling
•Ch.4.3: Reducing Branch Penalties with Dynamic
Hardware Prediction
•Ch.4.4: Taking Advantage of More ILP with
Multiple Issue
Exam
Details:
•Dec. 19th 1pm-4pm in Hamilton 717 (2:50 total time)
•About 7 large questions
•Mostly like homework questions & textbook
questions
•Open book, open notes
•No calculators / PDAs / etc.
Web
Page: The class
URL is: http://www.cs.columbia.edu/~jebara/3824
and will contain copies of handouts, homework assignments, and other
information. Course notes are available, come by my office. Solutions to
homeworks are available outside my office in CEPSR 605 in a black file tray.