← back

4541.775 Topics on Compilers Spring 2011

Syllabus Instructor: Bernhard Egger Office Hours: in 301-413 on Tuesdays, Thursdays, 09:30 - 11:30 Lecture: in 302-106 on Mondays, Wednesdays 11:00 - 12:15

News (5/23) There are no lecture slides for today's class. Ingoo will present his paper, and there is a reading assignment (see below). And please send me your preferences for the 3rd presentation block until Wednesday, May 25 (refer to the updated Syllabus) (5/2) There are no lecture slides for today's class. We will discuss the mid-term, plus there is a reading assignment till Wednesday (see below).

Lecture Material Syllabus and Schedule Introduction Dependence Dependence Testing Enhancing Parallelism Enhancing Fine-Grained Parallelism Coarse-Grained Parallelism VLIW Architectures: Introduction Instruction Scheduling on VLIW Architectures Introduction to CGRA CGRA Scheduling

Homework/Reading Assignments Assignment 1 (due: March 30, 2011, 11 a.m) (sample solution) Assignment 2 (due: April 25, 2011, 11 a.m) (sample solution) Reading Assignment (due: May 4, 2011): J.Fisher: Very Long Instruction Word architectures and the ELI­512 (ISCA'83) Reading Assignment (due: May 11, 2011): J.Fisher: Trace Scheduling: A Technique for Global Microcode Compaction (TC, Vol.30, No.7) Reading Assignment (due: June 1, 2011): B.Mei et al: Architecture exploration for a reconfigurable architecture template. (Design & Test of Computers, Vol 22, Iss. 2)

Student Presentations The Program Dependence Graph and its Use in Optimization (Honggyu Kim) Optimal Loop Parallelization (Ingoo Heo) Dependence Analysis for Pointer Variables (Gangwon Jo) Auto-vectorization of interleaved data for SIMD (Bertram Schmitt) Speculative Parallelization using State Separation and Multiple Value Predication (Christine Wagner) A Unified Modulo Scheduling and Register Allocation Technique for Clustered Processors (Honggyu Kim) Loop fusion for clustered VLIW architectures (Bertram Schmitt) Achieving Out-of-Order Performance with Almost In-Order Complexity (Gangwon Jo) Hybrid multithreading for VLIW processors (Christine Wagner) Design Space Exploration for Memory Subsystems of VLIW Architectures (Ingoo Heo) Architecture Exploration for a Reconfigurable Architecture Template (Ingoo Heo) Edge-centric Modulo Scheduling for Coarse-Grained Reconfigurable Architectures (Christine Wagner) Polymorphic Pipeline Array: A Flexible Multicore Accelerator with Virtualized Execution for Mobile Multimedia Applications (Bertram Schmitt) Resource Recycling: Putting Idle Resources to Work on a Composable Accelerator (Hongyu Kim) Placement-and-Routing-based Register Allocation for Coarse-Grained Reconfigurable Arrays (Gangwon Jo)