by Kurt Keutzer and Tim Mattson, Intel Corporation
Abstract
The key to writing high-quality parallel software is to develop a robust software design. This applies not only to the overall architecture of the program, but also to the lower layers in the software system where the concurrency and how it is expressed in the final program is [...]

[Read more →]

by Shameem Akhter and Jason Roberts, Intel Corporation
Abstract
Error diffusion in image and graphics processing programs represents an excellent example of a problem which at first may seem resistant to a multi-threaded approach, but upon further review will actually lend itself well to speed up possibilities. Using the Floyd-Steinberg algorithm for error diffusion, the trick lies [...]

[Read more →]

by Shameem Akhter and Jason Roberts, Intel Corporation
Abstract
The major CPU vendors are shifting gears, choosing to add parallelism support on-chip with multi-core processors in order to avoid many of the technological hurdles in boosting speeds, while still offering a better performing processor. However, if your software does not take advantage of these multiple cores, it [...]

[Read more →]

by Shameem Akhter and Jason Roberts, Intel Corporation
Abstract
This article describes the theory and practice of the principal parallel programming constructs that focus on threading and begins with the fundamental concepts of synchronization, critical section, and deadlock. It will discuss common issues associated with these concepts, as well as classical problems in the area of threading.

[Read more →]

by Shameem Akhter and Jason Roberts, Intel Corporation
Abstract
Parallel programming has been around for decades, though before the advent of multi-core processors, it was an esoteric discipline. Numerous programmers have tripped over the common stumbling blocks by now. By recognizing these problems you can avoid stumbling. Furthermore, it is important to understand the common problems before [...]

[Read more →]

by Shameem Akhter and Jason Roberts, Intel Corporation
Abstract
Parallel programming uses threads to enable multiple operations to proceed simultaneously. The entire concept of parallel programming centers on the design, development, and deployment of threads within an application and the coordination between threads and their respective operations. This article examines how to break up programming tasks into [...]

[Read more →]

by Shameem Akhter and Jason Roberts, Intel Corporation
Abstract
When programming for multiple thread or multiple core systems, it is important to realize the fundamental difference in memory allocation and access necessary from that of a single thread or core system. Contention and sharing with respect to memory locations can severely reduce the speedup gained from programming [...]

[Read more →]