Feature
| Benefit
|
Compiler and Debugger
|
- More Fortran 2003 Support
| Additional support includes ENUMERATOR, IEEE Floating Point Exception Handling, ALLOCATE extensions, array constructor changes and more to bring your Fortran apps closer to the standard. These join C interoperability features introduced in the last release to make it easier to develop mixed-language applications.
|
| OpenMP raises the parallelism abstraction away from the API, simplifying threading and making code more portable. Previously limited to loop-based data-parallelism, the new 3.0 standard simplifies both data and task parallelism.
|
| Take advantage of new Intel Streaming SIMD Extensions – automatically – through the compiler. No messy low-level coding to get the most from Intel® processors. Resetable for other hosts/targets.
|
| Supports your build by appropriately allocating files to available processors to take advantage of multicore processors and speed you through your edit/compile/debug cycle.
|
| More detailed optimization diagnostics for users who want to use our advanced optimizations to help the compiler do a better job of tuning their applications. |
| Find and analyze source file issues. Diagnostics include issues with OpenMP directives, boundary violations, memory corruptions, memory leak, buffer overflow and uninitialized memory. |
| Eclipse rich-client-platform based GUI makes it easier to see your application parallelism. Also offers command-line support. |
Intel® Math Kernel Library (Intel® MKL)
|
- New “layered” architecture
| The new architecture provides maximum support for different development environment configurations and processors in a single package.
|
| Link to the version of this layer that matches your development environment and rest assured that Intel MKL will not have threading incompatibilities with the threading in your application.
|
- Discrete Fourier Transform Interface
| The DftiCopyDescriptor function has been added for convenience when using the FFTs. The size of statically linked executables calling DFTI has been reduced significantly and complex storage is now available for real-to-real transforms.
|
| The capability to track and/or interrupt the progress of lengthy LAPACK computations has been added. A function called mkl_progress can be defined in a user application, which will be called regularly from a subset of the MKL LAPACK routines.
|
| With performance in mind, all VML functions are now threaded. And a new “Enhance Performance” mode is offered for applications where math-function inaccuracies don’t dominate parameter inaccuracies (e.g., Monte Carlo simulations and media applications).
|
| Improvements include threaded level-3 sparse BLAS triangular solvers and support for all data types (single precision, complex and double complex).
|
| New Linux Support | Fedora* 9, Ubuntu* 8.04, GNU tool chain 4.2 and 4.3. See the Release Notes for a complete list. |
Simplified installation
| Streamlined, simplified complete installation for a seamless one-step installation of all components.
|
New Online Support Community
| Our enhanced online community support forums and knowledge base search capabilities help you find answers more quickly. This is in addition to private, password-protected accounts available with Premier Support. Visit the Support Resources page to learn more!
|
Processor Support
| The addition of support for Intel® Atom™ processors continues to future-proof your investment with assurance of support for each successive generation of processors. That's a key advantage in a world where new hardware platforms come to market with awesome speed. For more details, see the Release Notes.
|
