Serving Eastern Massachusetts
| Course Name: | Introduction to Multicore Programming |
| Time & Date: | 7 - 9 PM, Tuesdays, March 30, April 6, 13, 20, 27 |
| Location: | Holiday Inn Select Hotel, 15 Middlesex Canal Park Rd, Woburn, MA |
| Speaker: | James T. Demers, Physical Sciences Inc. |
These lectures lay the foundation of multicore programming, covering the reasons for using it, challenges in system design, and software details. This begins by discussing platform-independent concepts, parallel programming constructs, and the decisions required for optimal use of available processing resources. A more in-depth presentation using the Cell BE processor as an example illustrates many of the fundamental ideas for performance improvement. Finally, a look at two competing concepts, lock-based and lock-free threads, shows the paths available as multicore programming matures.
There are no labs or required reading, but code will be examined in detail to illustrate the presented concepts.
Programming familiarity in C
Software developers using multiple processors, system engineers evaluating multicore solutions, and all engineers using newer processor technology.
Introduction to Multicore
Parallel Processing
Symmetric Multi-Processing
Asymmetric Multi-Processing
Valid Single-Core Concepts
Granularity
Performance Expectations
Benefits
Hurdles
Overview of SMP Resources
Program Co-ordination
Processes
Threads vs. Processes
Overview of AMP Resources
Available Software Packages
Asymmetric C Programming
Example: CUDA
Introduction to the Cell BE Architecture Computational Processing Considerations Data Movement Considerations Programming the PPE Programming the SPEs Performance Expectations
POSIX Thread Introduction
Thread Pitfalls
POSIX Thread Example: Banking with ATMs
Lock-free Overview
Contrasting Lock-free vs. Lock-based
Lock-free Example: Linked List
A graduate of MIT, James T. Demers has been developing parallel processing applications for 18 year at companies such as MIT Lincoln Laboratory, Mercury Computer Systems, and Physical Sciences. He has experience programming several types of processors, such as PowerPC, TI DSP, and the Cell BE, for image and signal processing applications. He can be reached at ieee@demers-family.org.
Payment received by March 16: IEEE Members $295
Payment received by March 16: Non-members $345
Payment received after March 16: IEEE Members $345
Payment received after March 16: Non-members $395