COMP 515: Advanced Compilation for Vector and Parallel Processors (Fall

...

2015)

Introduction

Are you curious about what makes a parallelizing compiler tick, and what program properties enable or prevent software from executing efficiently on vector and parallel computers?

...

Required Prerequisite: COMP 412 or equivalent

Lectures

NOTE: links Lectures marked ** are at non-standard times and locations.  Also, topics for future lectures will may be updated during the semester as the course progresses.

• Lecture 1 (08/27/2013, Profs. Palem & Sarkar25/2015): Compiler Challenges for High Performance Architectures (Chapter 1)  
• NO CLASS (08/27/2015): Rescheduled to 08/28/2015 below
• ** Lecture 2 (08/2928/2013, Prof. Palem2015, 4pm, DH 3076): Dependence: Theory and Practice (Chapter 2), Homework 1 assigned, due by 09/0508/20132015 
• Lecture 3 (09/03/2013, Prof. Palem01/2015): Dependence: Theory and Practice (Chapter 2, contd)
  
• Lecture 4 (09/05/2013, Prof. Palem03/2015): Dependence Testing (Chapter 3), Homework 1 due
• Lecture 5 (09/10/2013, Prof. Sarkar08/2015): Dependence Testing (Chapter 3, contd) Lecture 6 (09/12/2013, Prof. Sarkar): Dependence Testing (Chapter 3, contd), Homework ,  Homework 2 assigned, due by 09/2422/20132015
• No class NO CLASS (09/17/2013, self study): Prepare project proposal
• No class (09/19/2013, self study): Prepare project proposal
• 10/2015): Work on problem statement for your class projects
• Lecture 6 (09/15/2015Lecture 7 (09/24/2013, Prof. Sarkar): Dependence Testing (Chapter 3, contd),Preliminary  Preliminary Transformations (Chapter 4)
• Lecture 8 7 (09/26/2013, Prof. Sarkar17/2015): Preliminary Transformations (Chapter 4, contd), Homework 3 due by 10/08/2013Lecture 9 (10/01/2013, Prof. Sarkar Vectorization Algorithm (Chapter 2)
• Lecture 8 (09/22/2015):  Vectorization Algorithm (Chapter 2, contd), Enhancing Fine-Grained Parallelism (Chapter 5), Homework 3 assigned, due by 10/08/2015 
• Lecture 10 9 (1009/03/2013, Prof. Sarkar24/2015):  Enhancing  Enhancing Fine-Grained Parallelism (Chapter 5, contd)
• Lecture 11 10 (1009/08/2013, Prof. Sarkar): Unimodular Transformations 
• Lecture 12 (10/10/2013, Prof. Sarkar): Midterm Review, Exam 1 assigned, due by 10/17/2013
• Lecture 13 (10/21/2013, Prof. Sarkar): Polyhedral Compilation Framework, Exam 1 due
• 29/2015): Enhancing Fine-Grained Parallelism (Chapter 5, contd),  Coarse-Grained Parallelism (Chapter 6) 
• NO CLASS (10/01/2015): Work on class project
• Lecture 11 (10/06/2015): CoarseLecture 14 (10/22/2013, Prof. Palem): Coarse-Grained Parallelism (Chapter 6, contd)Lecture 15 , Optimal Weighted Loop Fusion for Parallel Programs
• NO CLASS (10/24/2013, Prof. Palem): Coarse-Grained Parallelism (Chapter 6, contd)
• Lecture 16 (10/29/2013, Prof. Palem): Coarse-Grained Parallelism (Chapter 6, contd), Control Dependence (Chapter 7)
• Lecture 17 (10/31/2013, Prof. Palem): Control Dependence (Chapter 7, contd)
• 08/2015): Homework 3 due, self-study, work on project
• NO CLASS (10/13/2015): MIDTERM RECESS (discuss project with instructor)
• Lecture 12 (10/15/2015): Midterm Review, Take-home Exam 1 assigned, due by 10/22/2015
• Lecture 13 (10/20/2015): Control Dependence (Chapter 7)
• NO CLASS (10/22/2015): Work on class project, Exam 1 due
• Lecture 14 (10/27/2015): Pipeline Parallelism, OpenMP 4.1 Doacross construct (guest lecture by Dr. Jun Shirako, Rice University)
• Lecture 15 (10/29/2015):  Control Lecture 18 (11/05/2013, Prof. Palem): Control Dependence (Chapter 7, contd), Compiler Improvement of Register Usage (Chapter 8) 
• Lecture 19 16 (11/07/2013, Prof. Palem03/2015): Compiler  Compiler Improvement of Register Usage (Chapter 8, contd), Homework 4 assigned, due by 11/24/2015
• Lecture 20 17 (11/12/2013, Prof. Palem): Compiler Improvement of Register Usage (Chapter 8, contd), 05/2015): Managing Cache (Chapter 9)
• Lecture 21 18 (11/14/2013, Prof. Palem10/2015): Managing  Managing Cache (Chapter 9, contd), ASTI Optimizer
• Lecture 22 19 (11/12/2015): Polyhedral Transformations (guest lecture by Prof. Uday Reddy, Indian Institute of Science)
• NO CLASS (11/17/2015): Self-study, work on project
• NO CLASS (11/19/2015): Self-study, work on project
• Lecture 20 (11/24/2015): Transformation Framework Case Studies
• NO CLASS (11/26/2015): THANKSGIVING
• Lecture 21 (12/01/2015): 19/2013, Prof. Palem): Interprocedural Analysis and Optimization (Chapter 11)Lecture 23 (11/21/2013, Prof. Palem): Compiling Array Assignments (Chapter 13)
• Lecture 24 22 (1112/26/2013, Prof. Sarkar03/2015):  EndEnd-Semester Summary (Chapters 7, 8, 9, 11, 13), Take-home Exam 2 assigned, due by 12/0911/20132015
• Lecture 25 23 (12/0308/20132015): Student project presentations Lecture 26 (12/05/2013): Student project presentations(special meeting time and place: 8:30am - 12noon, DH 3076)

References

1. Course text: Optimizing Compilers for Modern Architectures, Randy Allen & Ken Kennedy.
2. Automatic Selection of High Order Transformations in the IBM XL Fortran Compilers.  Vivek Sarkar. IBM Journal of Research and Development, 41(3), May 1997.
3. A General Framework for Iteration-Reordering Loop Transformations. Vivek Sarkar  and Radhika Thekkath. Proceedings of the ACM SIGPLAN 1992 Conference on  Programming Language Design and Implementation (PLDI), San Francisco, California,  pages 175-187, June 1992.
4. Optimal Weighted Loop Fusion for Parallel Programs. Nimrod Megiddo and Vivek  Sarkar. Proceedings of the Ninth Annual ACM Symposium on Parallel Algorithms  and Architecture (SPAA), Newport, Rhode Island, June, 1997, pages 282-291.
5. Array SSA form and its use in Parallelization. Kathleen Knobe and Vivek Sarkar.  Proceedings of the 25th ACM SIGPLAN-SIGACT Symposium on Principles of  Programming Languages, San Diego, California, January 1998.
   
   

...

Grading will be based on your performance on six four homeworks (weighted 25% 30% in all), one class project to be done in pairs (weighted 35%30%), and two exams (weighted 20% each).  Each project team will be expected to give a onean in-hour class presentation on their project at the end of the semester (Dec 3 or Dec 5), and to submit a report by December 1311, 20132015.

 The purpose of the homeworks is to train you to solve problems and to help deepen your understanding of concepts introduced in class. Homeworks are due on the dates and times specified in the course schedule. Homework is worth full credit when turned in on time. A 10% penalty per day will be levied on late homeworks, up to a maximum of 6 days. No submissions will be accepted more than 6 days after the due date.  If you believe we have made an error in grading your homework or exam, please bring the matter to our attention within one week after you receive the graded material.

...

Past offerings of COMP 515

• Fall 2013
• Fall 2011
• Spring 2009
• Spring 2006