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...

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

APPLICATION PERFORMANCE TOOLSThe /fast option maximizes speed across the entire program. For most Intel <strong>64</strong> <strong>and</strong><strong>IA</strong>-<strong>32</strong> processors, the “/fast” option is equivalent to “/O3 /Qipo /QxP” (-Q3 -ipo -static -xP on Linux). For Mac OS, the "-fast" option is equivalent to "-O3 -ipo".All the comm<strong>and</strong>-line options are described in Intel ® C++ Compiler documentation.A.1.1Recommended <strong>Optimization</strong> Settings for Intel <strong>64</strong> <strong>and</strong> <strong>IA</strong>-<strong>32</strong>Processors<strong>64</strong>-bit addressable code can only run in <strong>64</strong>-bit mode of processors that supportIntel <strong>64</strong> architecture. The optimal compiler settings for <strong>64</strong>-bit code generation isdifferent from <strong>32</strong>-bit code generation. Table A-1 lists recommended compiler optionsfor generating <strong>32</strong>-bit code for Intel <strong>64</strong> <strong>and</strong> <strong>IA</strong>-<strong>32</strong> processors. Table A-1 also appliesto code targeted to run in compatibility mode on an Intel <strong>64</strong> processor, but does notapply to running in <strong>64</strong>-bit mode. Table A-2 lists recommended compiler options forgenerating <strong>64</strong>-bit code for Intel <strong>64</strong> processors, it only applies to code target to run in<strong>64</strong>-bit mode. Intel compilers provide separate compiler binary to generate <strong>64</strong>-bitcode versus <strong>32</strong>-bit code. The <strong>64</strong>-bit compiler binary generates only <strong>64</strong>-bit addressablecode.Table A-1. Recommended <strong>IA</strong>-<strong>32</strong> Processor <strong>Optimization</strong> OptionsNeed Recommendation Comments• /QxT (-xT onLinux)Best performanceon Intel Core 2processor family <strong>and</strong>Intel Xeon processor3000 <strong>and</strong> 5100series, utilizingSSSE3 <strong>and</strong> otherprocessor-specificinstructionsBest performanceon Intel Core 2processor family <strong>and</strong>Intel Xeon processor3000 <strong>and</strong> 5100series, utilizingSSSE3; runs on non-Intel processorsupporting SSE2Best performanceon <strong>IA</strong>-<strong>32</strong> processorswith SSE3instruction support• /QaxT /QxW (-axT-xW on Linux)• Single code path• Will not run on earlier processors thatdo not support SSSE3• Multiple code path are generated• Be sure to validate your application onall systems where it may be deployed./QxP (-xP on Linux) • Single code path• Will not run on earlier processors.thatdo not support SSE3A-2

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

Saved successfully!

Ooh no, something went wrong!