Parallel Programming (T4-CHOICE)
The Parallel Programming course for the POWER Series, Onyx, and CHALLENGE workstations teaches programmers the techniques to write programs that take advantage of the parallel architecture of SGI's multiprocessor workstation. Students also learn how to optimize and tune existing programs to increase parallel performance.
Topics
This course covers these topics:
- Architecture Overview
- Performance Profiling Tools
- Parallel Optimization
- Memory Management for Parallel Processes
- IRIS Power C Accelerator (PCA)
- Power Fortran Accelerator (PFA)
- Use of the Multiprocessing Library
- Data and Function Partitioning
- Synchronization (Locks & Semaphores)
- Shared Memory
- Multi-processes Debugging
Objectives
Upon completion of this course, the student should be able to...
- Create parallel and concurrent programs using library and lower-level multiprocessing calls
- Use PFA and PCA to automatically parallelize existing code
Prerequisites
This course is intended for...
- Programmers who are familiar with parallel and concurrent programming applications
- Students who have Fortran or C programming skills
Note: The student is over-qualified if
he/she is already developing parallel and concurrent
programs using PFA or PCA, the Sequent library calls, the
IRIX arena library calls and SGI low-level primitives.
Course Details
Duration: 4.5 days
Format: Instructor-Led (Lecture/Lab)
Platforms Covered: Multiprocessor POWER Series, Onyx,
and CHALLENGE
Tuition :
Contact the Silicon Graphics Sales Office nearest you.
