GCC vs. LLVM/Clang On The AMD Richland APU

Written by Michael Larabel in Software on 6 July 2013 at 12:44 PM EDT. Page 1 of 4. 13 Comments.

Along with benchmarking the AMD A10-6800K "Richland" APU on Linux and its Radeon HD 8670D graphics, I provided some GCC compiler tuning benchmarks for this AMD APU with Piledriver cores. The latest Linux testing from the A10-6800K is a comparison of GCC 4.8.1 to LLVM/Clang 3.3 on this latest-generation AMD low-power system.

The GCC 4.8.1 vs/ LLVM Clang 3.3 compiler testing happened from the AMD A10-6800K when it was overclocked to 4.70GHz and running on the MSI FM2-A85XA-G65 motherboard. Ubuntu 13.10 was running on the system in its current development state with the Linux 3.9 kernel.

The Phoronix Test Suite automated benchmarking/testing Linux software was used with a variety of open-source C/C++ tests and the CFLAGS/CXXFLAGS were set for -O3 and -march=bdver2 during the benchmarking process.

AMD A10-6800K Compiler GCC vs. LLVM/Clang Performance

For those more curious about the Intel CPU performance with the different open-source compilers, see GCC 4.8 vs. LLVM Clang on Intel's Core i7 4770K Haswell from last month.

AMD A10-6800K Compiler GCC vs. LLVM/Clang Performance
AMD A10-6800K Compiler GCC vs. LLVM/Clang Performance

When getting things started for this AMD A10-6800K Richland compiler testing by using the MAFFT and BLAKE2 computational tests, GCC 4.8.1 was off with a slight advantage in performance over the recent LLVM/Clang 3.3 release.


Related Articles