Workshop on Duplicating, Deconstructing and Debunking
Schedule for Final Program
1:45pm-1:55pm Welcome and Introduction
1:55pm-3:30pm Session 1: Multiple Threads or Processors
Evaluation of Snoop-Energy Reduction Techniques for Chip-Multiprocessors
Speaker: Magnus Ekman, Chalmers University of Technology
Invited Response: Mikko Lipasti (also as proxy for Andreas Moshovos)
STiNG Revisited: Performance of Commercial Database Benchmarks on a CC-NUMA Computer System
Speaker: Russell Clapp, IBM
Performance Analysis of Simultaneous Multithreading in a PowerPC-based Processor
Speaker: Michel Hack, IBM
3:30pm-4:00pm Break w/refreshments
4:00pm-5:35pm Session 2: Caches and Branch Predictors
Evaluation of the Performance of Polynomial Set Index Functions
Speaker: Hans Vandierendonck, Ghent University
Invited Response: Antonio Gonzalez
The Benefit of Multiple Branch Prediction on Dynamically Scheduled Systems
Speaker: David M. Koppelman, Louisiana State University
Demystifying Intel Branch Predictors
Speaker: Milena Milenkovic, University of Alabama in Huntsville
5:35pm-5:50pm Short break
5:50pm-7:05pm Session 3: ILP Techniques
Revisiting Instruction Level Reuse
Speaker: Daniel Citron, IBM
Invited responses: Guri Sohi, Antonio Gonzalez
Memory Bypassing: Not Worth the Effort
Speaker: Gabriel H. Loh, Yale University
Invited response: Guri Sohi
7:05pm-7:15pm Conclusion and Evaluations