Ubuntu 12.04 LTS ARMv7 LLVM Clang 3.0 vs. GCC 4.6 Compiler PandaBoard ES Texas Instruments OMAP4460 ARMv7 dual-core ARM Cortex A9 1.2GHz benchmarking of Ubuntu 12.04 LTS hard-float with Linux 3.2 kernel. This ARMv7 compiler testing of the Precise packages for LLVM Clang 3.0 versus GCC 4.6.3 compilers. Same compiler flags used each time with -O3 CFLAGS/CXXFLAGS. Initial look at LLVM Clang C/C++ compiler performance under dual-core ARMv7 platform dual-core. ARM Linux benchmarking for a future article on Phoronix.com.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 1204218-SU-ARMCLANG163 ARM GCC 4.6.3 Processor: ARMv7 rev 10 @ 1.20GHz (2 Cores), Motherboard: OMAP4 Panda board, Memory: 904MB, Disk: 16GB SD16G, Graphics: Texas Instruments OMAP
OS: Ubuntu 12.04, Kernel: 3.2.0-1412-omap4 (armv7l), Desktop: Unity 2D 5.10.0, Display Server: X Server 1.11.3, Display Driver: fbdev 0.4.2, Compiler: GCC 4.6, File-System: ext4, Screen Resolution: 1920x1080
ARM LLVM-Clang 3.0 OS: Ubuntu 12.04, Kernel: 3.2.0-1412-omap4 (armv7l), Desktop: Unity 2D 5.10.0, Display Server: X Server 1.11.3, Display Driver: fbdev 0.4.2, Compiler: Clang 4.2.1, File-System: ext4, Screen Resolution: 1920x1080
TSCP OpenBenchmarking.org Nodes Per Second, More Is Better TSCP 1.81 AI Chess Performance ARM GCC 4.6.3 ARM LLVM-Clang 3.0 16K 32K 48K 64K 80K SE +/- 72.20, N = 5 SE +/- 56.82, N = 5 75346 63969
SciMark OpenBenchmarking.org Mflops, More Is Better SciMark 2.0 Computational Test: Composite ARM GCC 4.6.3 ARM LLVM-Clang 3.0 20 40 60 80 100 SE +/- 0.25, N = 4 SE +/- 0.05, N = 4 101.97 15.36
OpenBenchmarking.org Mflops, More Is Better SciMark 2.0 Computational Test: Monte Carlo ARM GCC 4.6.3 ARM LLVM-Clang 3.0 20 40 60 80 100 SE +/- 0.14, N = 4 SE +/- 0.02, N = 4 76.81 11.31
OpenBenchmarking.org Mflops, More Is Better SciMark 2.0 Computational Test: Fast Fourier Transform ARM GCC 4.6.3 ARM LLVM-Clang 3.0 4 8 12 16 20 SE +/- 0.26, N = 4 SE +/- 0.05, N = 4 15.67 5.12
OpenBenchmarking.org Mflops, More Is Better SciMark 2.0 Computational Test: Sparse Matrix Multiply ARM GCC 4.6.3 ARM LLVM-Clang 3.0 20 40 60 80 100 SE +/- 1.11, N = 4 SE +/- 0.09, N = 4 104.74 19.09
OpenBenchmarking.org Mflops, More Is Better SciMark 2.0 Computational Test: Dense LU Matrix Factorization ARM GCC 4.6.3 ARM LLVM-Clang 3.0 20 40 60 80 100 SE +/- 0.15, N = 4 SE +/- 0.04, N = 4 107.83 16.45
OpenBenchmarking.org Mflops, More Is Better SciMark 2.0 Computational Test: Jacobi Successive Over-Relaxation ARM GCC 4.6.3 ARM LLVM-Clang 3.0 40 80 120 160 200 SE +/- 0.34, N = 4 SE +/- 0.09, N = 4 204.81 24.83
OpenBenchmarking.org MB/s, More Is Better CacheBench Test: Write ARM GCC 4.6.3 ARM LLVM-Clang 3.0 600 1200 1800 2400 3000 SE +/- 47.53, N = 3 SE +/- 40.45, N = 3 2397.41 2987.02 1. (CC) gcc options: -lrt
OpenBenchmarking.org MB/s, More Is Better CacheBench Test: Read / Modify / Write ARM GCC 4.6.3 ARM LLVM-Clang 3.0 500 1000 1500 2000 2500 SE +/- 57.26, N = 6 SE +/- 2.61, N = 3 2376.22 416.29 1. (CC) gcc options: -lrt
FFTE FFTE is a package by Daisuke Takahashi to compute Discrete Fourier Transforms of 1-, 2- and 3- dimensional sequences of length (2^p)*(3^q)*(5^r). Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org MFLOPS, More Is Better FFTE 5.0 Test: N=64, 1D Complex FFT Routine ARM GCC 4.6.3 ARM LLVM-Clang 3.0 100 200 300 400 500 SE +/- 0.31, N = 3 SE +/- 1.11, N = 3 467.39 467.07 1. (F9X) gfortran options: -O3 -fomit-frame-pointer -fopenmp
N-Queens OpenBenchmarking.org Seconds, Fewer Is Better N-Queens 1.0 Elapsed Time ARM GCC 4.6.3 ARM LLVM-Clang 3.0 200 400 600 800 1000 SE +/- 0.99, N = 2 SE +/- 0.35, N = 2 821.79 705.09 1. (CC) gcc options: -static -O2
7-Zip Compression OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 9.20.1 Compress Speed Test ARM GCC 4.6.3 ARM LLVM-Clang 3.0 200 400 600 800 1000 SE +/- 10.40, N = 3 SE +/- 5.84, N = 3 859 844 1. (CXX) g++ options: -pipe -lpthread
C-Ray This is a test of C-Ray, a simple raytracer designed to test the floating-point CPU performance. This test is multi-threaded (16 threads per core), will shoot 8 rays per pixel for anti-aliasing, and will generate a 1600 x 1200 image. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better C-Ray 1.1 Total Time ARM GCC 4.6.3 ARM LLVM-Clang 3.0 1200 2400 3600 4800 6000 SE +/- 6.39, N = 3 SE +/- 63.86, N = 3 1170.77 5452.07 1. (CC) gcc options: -lm -lpthread -O3 -march=native
Smallpt Smallpt is a C++ global illumination renderer written in less than 100 lines of code. Global illumination is done via unbiased Monte Carlo path tracing and there is multi-threading support via the OpenMP library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Smallpt 1.0 Global Illumination Renderer; 100 Samples ARM GCC 4.6.3 130 260 390 520 650 SE +/- 0.33, N = 3 591 1. (CXX) g++ options: -fopenmp -O3 -march=native
OpenSSL OpenBenchmarking.org Signs Per Second, More Is Better OpenSSL 1.0.0e RSA 4096-bit Performance ARM GCC 4.6.3 ARM LLVM-Clang 3.0 1.035 2.07 3.105 4.14 5.175 SE +/- 0.00, N = 4 SE +/- 0.00, N = 4 4.60 4.60 1. (CC) gcc options: -O3 -lssl -lcrypto -ldl
ARM GCC 4.6.3 Processor: ARMv7 rev 10 @ 1.20GHz (2 Cores), Motherboard: OMAP4 Panda board, Memory: 904MB, Disk: 16GB SD16G, Graphics: Texas Instruments OMAP
OS: Ubuntu 12.04, Kernel: 3.2.0-1412-omap4 (armv7l), Desktop: Unity 2D 5.10.0, Display Server: X Server 1.11.3, Display Driver: fbdev 0.4.2, Compiler: GCC 4.6, File-System: ext4, Screen Resolution: 1920x1080
Testing initiated at 20 April 2012 08:57 by user phoronix.
ARM LLVM-Clang 3.0 Processor: ARMv7 rev 10 @ 1.20GHz (2 Cores), Motherboard: OMAP4 Panda board, Memory: 904MB, Disk: 16GB SD16G, Graphics: Texas Instruments OMAP
OS: Ubuntu 12.04, Kernel: 3.2.0-1412-omap4 (armv7l), Desktop: Unity 2D 5.10.0, Display Server: X Server 1.11.3, Display Driver: fbdev 0.4.2, Compiler: Clang 4.2.1, File-System: ext4, Screen Resolution: 1920x1080
Testing initiated at 20 April 2012 21:39 by user phoronix.