SPEC2000: Comparison of LLVM-2.9 and GCC4.6.1 on x86

The goal of these runs is to compare the performance of the latest versions of LLVM and GCC on x86 in 32-bit mode. I used the following options:

-O2 -mtune=generic and -O3 -ffast-math -mtune=generic for LLVM
-O1 -mtune=generic, -O2 -mtune=generic, -Ofast -mtune=generic, and -Ofast -ffast-math -flto -fwhole-program -mtune=generic for GCC

LLVM can not produce a correct code for 254.gap. Therefore -O0 was used for both LLVM and GCC for 254.gap.

The used machine is 3.4Ghz Intel Core i7-2600 with 8GB memory under Fedora Core14.

GCC-4.6.1 was configured with --enable-checking=release and was bootstrapped which means that it was compiled by itself. LLVM-2.9 x86 binaries are taken from LLVM download site. All changes are given relative to LLVM-2.9 run with -O2 -mtune=generic.

Compilation Speed
Code Size

Here the code size is the corresponding text and data segment size.
- SPECInt2000
- SPECFP2000
Generated Code Performace
- SPECInt2000
- SPECFP2000

Last modified: 09/08/2011 - vmakarov at redhat dot com

Return to index page.

The used machine is 3.4Ghz Intel Core i7-2600 with 8GB memory under Fedora Core14.

Compilation Speed

Code Size

SPECInt2000

SPECFP2000

Generated Code Performace

SPECInt2000

SPECFP2000