As I continue to explore different Ultrabook capabilities, in this blog I decided to look into a powerful threading and performance optimization tool for C/C++, .NET, and FORTRAN developers who need to understand an application's serial and parallel behavior to improve performance and scalability: Intel® VTune™ Amplifier XE 2011.
In all my years in college I have used a laptop PC. Recently, I’ve been tempted to get a MacBook Air Pro because of its light-weight and stylish look. But, since starting at Intel, I’ve been able to explore and play with an Ultrabook and it has changed my mind completely. The reason is simple: Ivy Bridge Ultrabooks are just amazing! Ultrabooks will be even more amazing when you add the touch screen capability of Window 8 later this year.
Some people say that extending Moore’s Law into the future isn’t necessary, and that today’s computer hardware and software is good enough. This is a dubious notion given the history of the information technology industry. All too often, statements about good-enough computing capabilities, or innovations that will never find a market in the first place, are proven wrong – no matter who’s making the claim.
Apply threading to data-decomposition problems in the Implementation Phase of application development. Data-decomposition problems are situations where multiple threads need to be assigned to perform the same functionality on different data.
The serial code shown below computes the value of pi.
In the first part of this article I described how you can “have your cake and eat it too” with respect to programmable use of Hyper Threading or no Hyper Threading through use of thread team selection attributes on the parallel_for construct in the QuickThread® programming tool kit.
In this part I will describe the test bed application and results data as run on an Intel Core i7 2600K Sandy Bridge (no over clocking).
Open Parallel is a research and development company that focuses on parallel programming and multicore development. We are a bunch of highly skilled geeks from various backgrounds that work together on problems in parallel programming and software development for multicore and manycore platforms.
- Page 1