The PARSEC Benchmark Suite Tutorial

Abstract

This tutorial introduces its attendants to the Princeton Application Repository for Shared-Memory Computers (PARSEC), a benchmark suite for studies of Chip-Multiprocessors (CMPs). Previous available benchmarks for multiprocessors have focused on high-performance computing applications and used a limited number of synchronization methods. PARSEC includes emerging applications in recognition, mining and synthesis (RMS) as well as systems applications which mimic large-scale multi-threaded commercial programs.

The tutorial covers three areas: First, the PARSEC benchmark programs and their runtime properties are discussed. Several new features are introduced. Second, participants will learn how to use the PARSEC framework and its tools to build, run and instrument PARSEC programs efficiently. Three, the ongoing work for benchmarking network workloads is discussed, which calls for contributions on providing network workloads to us.

Target Audience

The tutorial is aimed at researchers in computer architecture and related fields who use benchmark programs. It covers aspects relevant to performance measurement with PARSEC.

Schedule

The tutorial will take place on Saturday morning, June 4.

09:00am - 10:00am   Part 1: Introduction to PARSEC & PARSEC Workloads
10:00am - 10:30am   Break
10:30am - 11:30pm   Part 2: Working with PARSEC
11:30am - 11:45am   Break
11:45am - 12:30pm   Part 3: Ongoing Work, Call for Contributions & Further Discussion


Tutorial sildes are avaliable for download: [pdf]

Kindly Remind:
Breakfast is available only from 8:00-9:00 AM.
Morning coffee is only available from 9:45-10:30 AM.
Lunch is from 12:30-1:30.