LLVM/Clang 3.3 Performing Against GCC For Old Intel CPU

Written by Michael Larabel in Software on 23 April 2013 at 08:33 PM EDT. Page 1 of 3. 5 Comments.

Generally when delivering new Linux compiler benchmarks on Phoronix it's from x86/ARM hardware within the past two years. It's the most recent generations of hardware that excites us the most and generally where the professional Linux software developers are focusing their time and resources. However, after seeing the recent LLVM/Clang 3.3 performance improvements for this forthcoming open-source compiler release, we decided to go back a bit in CPU history.

Rather than benchmarking the very latest Intel Ivy/Sandy Bridge or AMD Bulldozer 2 hardware or an ARM Cortex-A15, in this article an Intel Core 2 Duo is being used for some new GCC vs. LLVM/Clang benchmarks. After having an old Apple Mac Mini out of storage for delivering the Intel i915 Gallium3D benchmarks, it was tasked by the Phoronix Test Suite in conjunction with OpenBenchmarking.org for doing some new compiler benchmarks.

This Mac Mini sports an Intel Core 2 Duo T5600 processor that is a dual-core part clocked at 1.83GHz. The Intel T5600 is a 64-bit "Merom" processor from 2006 that supports SSE3/SSE4. This x86_64 dual-core system had 1GB of RAM, 80GB of HDD-backed storage, and was running Ubuntu 13.04 with the Linux 3.8 kernel. Being benchmarked for this older CPU compiler comparison were GCC 4.7.3, LLVM/Clang 3.2, and LLVM/Clang 3.3 SVN as of this week. The same compiler configuration was used for each tested compiler.

Let's see where things stand when benchmarking GCC and LLVM/Clang on this nearly seven-year-old processor.


Related Articles