Hybrid Programming for Shared-Memory and Clustered SMP Systems

September 8 - 11, 2008

Pittsburgh Supercomputing Center
300 S. Craig Street, Pittsburgh, PA

Abstract

This workshop will present programming models and techniques for writing efficient parallel code on contemporary and future supercomputers with extensive shared memory, or hierarchical architectures with smaller shared-memory components. Two important examples of systems to which these techniques apply are the SGI Altix and networked clusters of multicore processors. Expert instructors from PSC and SGI will review MPI, OpenMP, and hardware architecture prior to launching into detailed treatments of programming for hybrid parallelism, performance analysis, and optimization. This is a "bring your own code" workshop. Participants are encouraged to bring an application to focus on during the hands-on sessions to maximize the workshop's effectiveness. Examples will be provided for participants who cannot bring a research code. Experienced PSC computational scientists will provide support regarding the topics covered, including hybrid algorithms and implementation strategies and performance engineering.

Agenda

Monday, September 8, 2008
9:00	Welcome/Orientation	Urbanic
9:15	Intro to PSC	Urbanic
9:30	Intro to Parallel Computing	Urbanic
10:30	Break
10:45	SGI Altix Architecture	Cisneros
12:00	Lunch
1:00	MPI Basics	Urbanic
2:00	Break
2:15	Hands-on Exercises
4:00	Altix Memory and Time Management	Cisneros

Tuesday, September 9, 2008
9:00	Scalable Coding: Laplace Case Study	Urbanic
10:00	Hands-on Exercises
10:30	Break
10:45	Hands-on Exercises
12:00	Lunch
1:00	OpenMP Basics	Blood
2:00	Break
2:15	Compiling for Shared-Memory Parallelism	Cisneros
3:15	Hands-on Exercises

Wednesday, September 10, 2008
9:00	Intro to Final Hybrid Exercise	Blood
9:30	Hands-On Exercises
10:30	Break
10:45	Hands-on Exercises
12:00	Lunch
1:00	Debugging	Cisneros
2:00	Break
2:15	Hands-On Exercises
4:00	Tuning Parallel Applications	Cisneros

Thursday, September 11, 2008
9:00	Performance Analysis	Cisneros
10:00	Single-CPU Optimization Techniques	Cisneros
11:00	Break
11:15	Hands-on Exercises
12:00	Lunch
1:00	Outro to Parallel Computing	Urbanic
2:00	Break
2:15	Hands-On Exercises
3:00	SGI Altix Systems I/O Usage and Performance	Cisneros
3:45	Departure Survey and Hands-On Exercises

Registration

Accommodations

A block of rooms is reserved at the

Holiday Inn Select - University Center 100 Lytton Avenue Pittsburgh, PA 15213 Phone: 412-682-6200 Fax: 412-681-4749 Group Name: Hybrid Programming for Multicore Systems Rate: $108 + tax/night Reservation Cut-off date: August 18, 2008