26.11.2012 Views

Compiler Usage Guidelines for 64-Bit Operating Systems on AMD64 ...

Compiler Usage Guidelines for 64-Bit Operating Systems on AMD64 ...

Compiler Usage Guidelines for 64-Bit Operating Systems on AMD64 ...

SHOW MORE
SHOW LESS

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

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

32035 Rev. 3.22 November 2007<br />

5.1.2 Peak Command-line Opti<strong>on</strong>s<br />

<str<strong>on</strong>g>Compiler</str<strong>on</strong>g> <str<strong>on</strong>g>Usage</str<strong>on</strong>g> <str<strong>on</strong>g>Guidelines</str<strong>on</strong>g> <str<strong>on</strong>g>for</str<strong>on</strong>g> AMD<str<strong>on</strong>g>64</str<strong>on</strong>g> Plat<str<strong>on</strong>g>for</str<strong>on</strong>g>ms<br />

The table below specifies the best-known peak switches <str<strong>on</strong>g>for</str<strong>on</strong>g> various benchmarks in the SPECcpu2006<br />

suite <str<strong>on</strong>g>for</str<strong>on</strong>g> the <str<strong>on</strong>g>64</str<strong>on</strong>g>-bit PGI Release 7.1 compilers <str<strong>on</strong>g>for</str<strong>on</strong>g> Linux® <strong>on</strong> AMD Athl<strong>on</strong> <str<strong>on</strong>g>64</str<strong>on</strong>g> processor based<br />

plat<str<strong>on</strong>g>for</str<strong>on</strong>g>ms and AMD Opter<strong>on</strong> processor-based plat<str<strong>on</strong>g>for</str<strong>on</strong>g>ms.<br />

Table 10. Best-Known Peak Switches <str<strong>on</strong>g>for</str<strong>on</strong>g> the <str<strong>on</strong>g>64</str<strong>on</strong>g>-<str<strong>on</strong>g>Bit</str<strong>on</strong>g> PGI <str<strong>on</strong>g>Compiler</str<strong>on</strong>g>s <str<strong>on</strong>g>for</str<strong>on</strong>g> Linux ®<br />

Applicati<strong>on</strong> Area Benchmark Language Best Known Peak Switches<br />

Programming<br />

Language<br />

400.perlbench ANSI C<br />

Compressi<strong>on</strong> 401.bzip2 ANSI C<br />

GNU C compiler 403.gcc C<br />

Combinati<strong>on</strong>al<br />

Optimizati<strong>on</strong><br />

Artificial<br />

Intelligence: Go<br />

Search Gene<br />

Sequence<br />

Artificial<br />

Intelligence:<br />

Chess<br />

Physics / Quantum<br />

Computing<br />

Video<br />

compressi<strong>on</strong><br />

Discrete Event<br />

Simulati<strong>on</strong><br />

Path-finding<br />

Algorithms<br />

CINT2006 3<br />

pgcc -w -fast -O4 -Mfprelaxed -Msmartalloc=huge:448<br />

-Mnounroll -Mpfi(pass 1) -Mpfo(pass 2)<br />

-Mipa=inline(pass 2) -tp barcel<strong>on</strong>a-<str<strong>on</strong>g>64</str<strong>on</strong>g><br />

-DSPEC_CPU_LP<str<strong>on</strong>g>64</str<strong>on</strong>g> -DSPEC_CPU_LINUX_X<str<strong>on</strong>g>64</str<strong>on</strong>g><br />

pgcc -w -fast -O4 -Msmartalloc=huge:448<br />

-Mpfi(pass 1) -Mpfo(pass 2) -tp barcel<strong>on</strong>a-<str<strong>on</strong>g>64</str<strong>on</strong>g> -Mpfo<br />

-DSPEC_CPU_LP<str<strong>on</strong>g>64</str<strong>on</strong>g><br />

pgcc -w -fastsse -Mfprelaxed -Msmartalloc=huge:448<br />

-Mipa=fast, inline -tp barcel<strong>on</strong>a<br />

429.mcf ANSI C 1 pgcc -w -fastsse -Mipa=fast, inline:1<br />

-Msmartalloc=huge:420 -tp barcel<strong>on</strong>a<br />

445.gobmk C<br />

456.hmmer C<br />

458.sjeng ANSI C<br />

462.libquantum “C99”<br />

pgcc -w -fast -O4 -Msmartalloc=huge:448 -Mfprelaxed<br />

-Mnovect -tp barcel<strong>on</strong>a-<str<strong>on</strong>g>64</str<strong>on</strong>g> -Mpfi(pass 1)<br />

-Mpfo(pass 2) -Mipa=fast(pass 2)<br />

-DSPEC_CPU_LP<str<strong>on</strong>g>64</str<strong>on</strong>g><br />

pgcc -w -fast -Msmartalloc=huge:448 -Mfprelaxed<br />

-Msafeptr -Mipa=c<strong>on</strong>st, ptr, arg -tp barcel<strong>on</strong>a-<str<strong>on</strong>g>64</str<strong>on</strong>g><br />

-DSPEC_CPU_LP<str<strong>on</strong>g>64</str<strong>on</strong>g><br />

pgcc -w -fast -Msmartalloc=huge:448 -Mfprelaxed<br />

-tp barcel<strong>on</strong>a-<str<strong>on</strong>g>64</str<strong>on</strong>g> -Mpfi(pass 1) -Mpfo(pass 2)<br />

-Mipa=fast, inline:1, noarg(pass 2)<br />

-DSPEC_CPU_LP<str<strong>on</strong>g>64</str<strong>on</strong>g><br />

pgcc -w -fast -Mfprelaxed -Msmartalloc=huge:448<br />

-Munroll=m:4 -Mipa=fast, inline, noarg<br />

-DSPEC_CPU_LP<str<strong>on</strong>g>64</str<strong>on</strong>g> -DSPEC_CPU_LINUX<br />

4<str<strong>on</strong>g>64</str<strong>on</strong>g>.h2<str<strong>on</strong>g>64</str<strong>on</strong>g>ref C Use base binaries and/or base results <str<strong>on</strong>g>for</str<strong>on</strong>g> peak.<br />

471.omnetpp C++ Use base binaries and/or base results <str<strong>on</strong>g>for</str<strong>on</strong>g> peak.<br />

473.astar C++ Use base binaries and/or base results <str<strong>on</strong>g>for</str<strong>on</strong>g> peak.<br />

Notes:<br />

1. Mathematical library (libm) required<br />

2. Boost Library required<br />

3. Smartheap libraries utilized. If the Smartheap libraries are not loaded, xalancbmk per<str<strong>on</strong>g>for</str<strong>on</strong>g>ms better with the<br />

-Msmartalloc=huge:160 opti<strong>on</strong>.<br />

Chapter 5 Peak Opti<strong>on</strong>s <str<strong>on</strong>g>for</str<strong>on</strong>g> SPEC ® -CPU Benchmark Programs 57

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

Saved successfully!

Ooh no, something went wrong!