Guru in a Box—For Win32
One of the tool's most powerful features is an artificial-intelligence–like feature that offers advice for fixing pesky source code. With version 6.1, this feature is now called Intel Tuning Assistant, and although it behaves slightly differently than the Code Coach advisor in version 5 (Intel Tuning Assistant is integrated with the source code viewer, rather than being its own special viewer), its advice is just as valuable.
When it comes to building .NET applications, there's only one problem: The Intel Tuning Assistant is for natively compiled languages only, such as Intel's own C++ or Fortran compilers, Microsoft Visual C++ 6.0, or Borland C++ Builder. While Intel Tuning Assistant can also work with some Java runtimes under Windows, the code advisor does not yet work with applications that execute through the .NET Framework's Common Language Runtime, or on applications on Linux, the other platform that Intel VTune Performance Analyzer supports. (Intel VTune Performance Analyzer supports Linux remotely, meaning that you can instrument and capture data on Linux applications running on kernel 2.4.2 or later, but all post-runtime performance analysis must take place on a Windows workstation.)
That's not to say that a developer couldn't or shouldn't use the Intel VTune Performance Analyzer with .NET. You should: thanks to its integration with Visual Studio .NET, and the power and simplicity of its Source Viewer and Call Graph analysis, you'll gain a tremendous understanding of your application's runtime profile. And that, in turn, will help you ensure that the applications run faster and scale more efficiently.
But don't take my word for it. Check out the Intel VTune Performance Analyzer yourself, using a 30 day evaluation copy you can download from http://www.intel.com/software/products/global/eval.htm
Related Links
Intel VTune Performance Analyzer 6.0 Beats Bottlenecks in Code
Boosting the Performance of Microsoft .NET
Developing Multi-Threaded Applications Using .NET
The .NET Platform Developer Center
About The Author
A former mainframe software developer and systems analyst, Alan Zeichick is principal analyst at Camden Associates, an independent technology research firm focusing on networking, storage, and software development. He can be reached at zeichick@camdenassociates.com.
The information, opinions, and recommendations in this column are provided by the author, Alan Zeichick. Intel, Tapestry.net, and their subsidiaries do not necessarily endorse or represent the accuracy of the author's information opinions or recommendations and any reliance upon the author's statements is solely at your own risk.
To get a pdf version of this article, click download, below.
© 2004 Intel Corp.
 | J2EE performance optimization, part 1: laying the foundation by Kingsum Chow, PhD, senior performance architect with the Managed ... |
If you're interested in this topic, these articles may be helpful:
| Intel® Integrated Performance Primitives 4.1 Intel® Integrated Performance Primitives (Intel® IPP) is a library o... |
| Threading Games for High Performance on Intel® Processors The evolution of the multi-threaded processor design is the trend for ... |
| EJB best practices: the fine points of data validation - how to get the best performance out of your validation code by Brett McLaughlin, author and editor, O'Reilly and Associates.First ... |
| High Performance Linux Clusters with OSCAR, Rocks, OpenMosix, and MPI by Joseph D. Sloan. O'Reilly Media Inc. To the outside world, a "su... |
| Data access performance in ADO.NET by Ramesh Theivendran, architect, Borland Software Corp. First publish... |
Related Jobs:
