SC12 Home > SC12 Schedule > SC12 Presentation - Optimizing pF3D using Model-Based, Dynamic Parallelism

SCHEDULE: NOV 10-16, 2012

When viewing the Technical Program schedule, on the far righthand side is a column labeled "PLANNER." Use this planner to build your own schedule. Once you select an event and want to add it to your personal schedule, just click on the calendar icon of your choice (outlook calendar, ical calendar or google calendar) and that event will be stored there. As you select events in this manner, you will have your own schedule to guide you through the week.

Optimizing pF3D using Model-Based, Dynamic Parallelism

SESSION: ACM Student Research Competition Poster Session

EVENT TYPE: ACM Student Research Competition

TIME: 5:15PM - 7:00PM

SESSION CHAIR: Torsten Hoefler


ROOM:East Entrance

Optimizing parallel applications for performance and power in current and future systems poses significant challenges. A single node today present multiple levels of parallelism including multiple SMT-threads, cores, sockets, and memory domains. Determining the optimal concurrency and mapping of an application to the underlying processing units may be intractable for online optimization and challenging for efficient offline search. In this work, we present a framework to dynamically optimize the performance of parallel programs based on model predictions of the optimal configuration. We optimize the performance of kernels from pF3D, a real-world multi-physics code used to simulate the interaction between a high-intensity laser and a plasma. Our results show that our approach predicts performance within 6% of the optimal in average and achieve performance improvements from 1.03x to 5.95x compared to the Linux default setting.

Chair/Author Details:

Torsten Hoefler (Chair) - ETH Zurich

ChunYi Su - Virginia Tech

Add to iCal  Click here to download .ics calendar file

Add to Outlook  Click here to download .vcs calendar file

Add to Google Calendarss  Click here to add event to your Google Calendar