CSE 533
Advanced Computer Architectures
( Spring 2008 )

Announcements

The Open-Book-Final Exam is to be held on 2^nd of Jun, 2008, 14:00-17:00.
The project is due 2^nd of Jun, 2008.Click Here
Presentations:
Dataflow Architectures 6pp
Multiclustered Architectures 6pp
SMT Architectures 6pp
VLIW and EPIC Architectures 6pp
Cell Architecture 6pp
CMP Architectures 6pp
Quantum Architectures 6pp

General Information

Instructor: Assist.Prof. Gurhan Kucuk, A-409, x1416 E-Mail: gkucuk (at) cse.yeditepe.edu.tr URL: http://cse.yeditepe.edu.tr/~gkucuk Lecture Hrs: Tue 14:00-17:00, Room: A-412 Office Hrs: Tue 11:00-11:50, 13:00-13:50 and by appointment


Course Syllabus Description: Fundamentals of computer design, instruction set principles, pipelining and pipeline hazards, instruction level parallelism, overcoming pipeline hazards, superscalar, superpipelined processors, memory-hierarchy design, VLIW, CMP, SMT, Dataflow, Multicluster architectures Contents: Fundamentals of Computer Design Technology and Computer Usage Trends Measuring and Reporting Performance Why Worry About Power? Fallacies and Pitfalls Pipelining Basic Principles Pipeline Hazards Handling Multicycle Operations Advanced Pipelining Instruction Level Parallelism (ILP) Compiler Support for Exploiting ILP Dynamic Scheduling Reducing Branch Penalties More ILP with Multiple Issue (In-Order vs. Out-of-Order Execution Core) Memory-Hierarchy Design The ABCs of Caches Reducing Cache Misses Reducing Cache Miss Penalty Reducing Hit Time Main Memory Virtual Memory Other Directions: VLIW, SMT, CMP, Dataflow and Multicluster Machines Overview of Parallel Architectures Grading: 1 Midterm Exam: %20 (8^th April 2008) Final exam: %30 Programming Projects: %10 Paper Presentation: %10 Homeworks & Quizzes: %10 Final Project: %20 Books: John L. Hennessy and David A. Patterson Computer Architecture: A Quantitative Approach Second Edition Morgan Kaufman, 2002. J. Shen and M. Lipasti Modern Processor Design: Fundamentals of Superscalar Processors McGraw-Hill, 2004 Lecture Notes 1-Introduction 1 slide pp 6 slides pp 2a-Performance, The Iron Law, The Amdahl's Law 1 slide pp 6 slides pp 2b-Power - I 1 slide pp 6 slides pp, Supplementary Slides Power - II, Power - III 3-ISA 1 slide pp 6 slides pp 4-Pipelining 1 slide pp 6 slides pp Supplementary Notes for Pipelining, Another Set of Slides 5-Static Scheduling 1 slide pp 6 slides pp 6-Dynamic Scheduling 1 slide pp 6 slides pp , Supplementary Slides 1 slide pp 6 slides pp 7-Register Renaming 1 slide pp 6 slides pp 8-Register Renaming with Re-Order Buffer 1 slide pp 6 slides pp 9-Branch Prediction 1 slide pp Supplementary 10-Advanced Pipelines 2 slides pp 11-Memory Hierarchy Design 1 slide pp 6 slides pp 12-Virtual Memory 1 slide pp 6 slides pp 13-SRAM vs. DRAM 1 slide pp 6 slides pp Old Presentations: 1-Chip MultiProcessors 1 slide pp 6 slides pp 2-Load Value Speculation 1 slide pp 6 slides pp 3-SMT processors 1 slide pp 6 slides pp 4-Out-of-Order Commit1 slide pp 6 slides pp 5-Runahead Execution1 slide pp 6 slides pp Resources Download M-Sim Download Cygwin SimpleScalar Technical Report Assignments Second Homework Assignment. Due date: 8^th of April, 2008. Click Here Grades