13.07.2015 Views

Intel® 64 and IA-32 Architectures Optimization Reference Manual

Intel® 64 and IA-32 Architectures Optimization Reference Manual

Intel® 64 and IA-32 Architectures Optimization Reference Manual

SHOW MORE
SHOW LESS
  • No tags were found...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

APPLICATION PERFORMANCE TOOLSlibrary include linear algebra consisting of LAPACK <strong>and</strong> BLAS, Discrete FourierTransforms (DFT), vector transcendental functions (vector math library/VML) <strong>and</strong>vector statistical functions (VSL). Intel MKL is optimized for the latest features<strong>and</strong> capabilities of the Intel Pentium 4 processor, Pentium M processor, Intel Xeonprocessors <strong>and</strong> Intel ® Itanium ® 2 processors.• Intel ® Integrated Performance Primitives for Linux* <strong>and</strong> Windows*: IPP is across-platform software library which provides a range of library functions forvideo decode/encode, audio decode/encode, image color conversion, computervision, data compression, string processing, signal processing, image processing,JPEG decode/encode, speech recognition, speech decode/encode, cryptographyplus math support routines for such processing capabilities.Intel IPP is optimized for the broad range of Intel microprocessors: Intel Core 2processor family, Dual-core Intel Xeon processors, Intel Pentium 4 processor,Pentium M processor, Intel Xeon processors, the Intel Itanium architecture,Intel ® SA-1110 <strong>and</strong> Intel ® PCA application processors based on the IntelXScale ® microarchitecture. With a single API across the range of platforms, theusers can have platform compatibility <strong>and</strong> reduced cost of development.A.3.1Benefits SummaryThe overall benefits the libraries provide to the application developers are as follows:• Time-to-Market — Low-level building block functions that support rapidapplication development, improving time to market.• Performance — Highly-optimized routines with a C interface that giveAssembly-level performance in a C/C++ development environment (MKL alsosupports a Fortran interface).• Platform tuned — Processor-specific optimizations that yield the bestperformance for each Intel processor.• Compatibility — Processor-specific optimizations with a single applicationprogramming interface (API) to reduce development costs while providingoptimum performance.• Threaded application support — Applications can be threaded with theassurance that the MKL <strong>and</strong> IPP functions are safe for use in a threadedenvironment.A.3.2<strong>Optimization</strong>s with the Intel ® Performance LibrariesThe Intel Performance Libraries implement a number of optimizations that arediscussed throughout this manual. Examples include architecture-specific tuningsuch as loop unrolling, instruction pairing <strong>and</strong> scheduling; <strong>and</strong> memory managementwith explicit <strong>and</strong> implicit data prefetching <strong>and</strong> cache tuning.A-13

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!