Fedora 21 GCC Compiler Optimizations Compiler optimization benchmarks by Michael Larabel for a future article on Phoronix.com. -O0: Processor: Intel Xeon E5-1680 v3 @ 3.80GHz (16 Cores), Motherboard: MSI X99S SLI PLUS (MS-7885) v1.0, Chipset: Intel Xeon E5 v3/Core, Memory: 16384MB, Disk: 80GB INTEL SSDSCKGW08, Graphics: eVGA NVIDIA GeForce GT 740 1024MB, Audio: Intel C610/X99, Monitor: VE228, Network: Intel Connection OS: Fedora 21, Kernel: 3.18.0-0.rc3.git2.2.fc22.1.x86_64 (x86_64), Desktop: GNOME Shell 3.14.1.5, Display Server: X Server 1.16.1, Display Driver: nouveau 1.0.10, OpenGL: 3.3 Mesa 10.3.2 Gallium 0.4, Compiler: GCC 4.9.2 20141101 + LLVM 3.4.2, File-System: ext4, Screen Resolution: 1920x1080 -O2: Processor: Intel Xeon E5-1680 v3 @ 3.80GHz (16 Cores), Motherboard: MSI X99S SLI PLUS (MS-7885) v1.0, Chipset: Intel Xeon E5 v3/Core, Memory: 16384MB, Disk: 80GB INTEL SSDSCKGW08, Graphics: eVGA NVIDIA GeForce GT 740 1024MB, Audio: Intel C610/X99, Monitor: VE228, Network: Intel Connection OS: Fedora 21, Kernel: 3.18.0-0.rc3.git2.2.fc22.1.x86_64 (x86_64), Desktop: GNOME Shell 3.14.1.5, Display Server: X Server 1.16.1, Display Driver: nouveau 1.0.10, OpenGL: 3.3 Mesa 10.3.2 Gallium 0.4, Compiler: GCC 4.9.2 20141101 + LLVM 3.4.2, File-System: ext4, Screen Resolution: 1920x1080 -O3: Processor: Intel Xeon E5-1680 v3 @ 3.80GHz (16 Cores), Motherboard: MSI X99S SLI PLUS (MS-7885) v1.0, Chipset: Intel Xeon E5 v3/Core, Memory: 16384MB, Disk: 80GB INTEL SSDSCKGW08, Graphics: eVGA NVIDIA GeForce GT 740 1024MB, Audio: Intel C610/X99, Monitor: VE228, Network: Intel Connection OS: Fedora 21, Kernel: 3.18.0-0.rc3.git2.2.fc22.1.x86_64 (x86_64), Desktop: GNOME Shell 3.14.1.5, Display Server: X Server 1.16.1, Display Driver: nouveau 1.0.10, OpenGL: 3.3 Mesa 10.3.2 Gallium 0.4, Compiler: GCC 4.9.2 20141101 + LLVM 3.4.2, File-System: ext4, Screen Resolution: 1920x1080 -O3 -march=native: Processor: Intel Xeon E5-1680 v3 @ 3.80GHz (16 Cores), Motherboard: MSI X99S SLI PLUS (MS-7885) v1.0, Chipset: Intel Xeon E5 v3/Core, Memory: 16384MB, Disk: 80GB INTEL SSDSCKGW08, Graphics: eVGA NVIDIA GeForce GT 740 1024MB, Audio: Intel C610/X99, Monitor: VE228, Network: Intel Connection OS: Fedora 21, Kernel: 3.18.0-0.rc3.git2.2.fc22.1.x86_64 (x86_64), Desktop: GNOME Shell 3.14.1.5, Display Server: X Server 1.16.1, Display Driver: nouveau 1.0.10, OpenGL: 3.3 Mesa 10.3.2 Gallium 0.4, Compiler: GCC 4.9.2 20141101 + LLVM 3.4.2, File-System: ext4, Screen Resolution: 1920x1080 -O3 -march=native -flto: Processor: Intel Xeon E5-1680 v3 @ 3.80GHz (16 Cores), Motherboard: MSI X99S SLI PLUS (MS-7885) v1.0, Chipset: Intel Xeon E5 v3/Core, Memory: 16384MB, Disk: 80GB INTEL SSDSCKGW08, Graphics: eVGA NVIDIA GeForce GT 740 1024MB, Audio: Intel C610/X99, Monitor: VE228, Network: Intel Connection OS: Fedora 21, Kernel: 3.18.0-0.rc3.git2.2.fc22.1.x86_64 (x86_64), Desktop: GNOME Shell 3.14.1.5, Display Server: X Server 1.16.1, Display Driver: nouveau 1.0.10, OpenGL: 3.3 Mesa 10.3.2 Gallium 0.4, Compiler: GCC 4.9.2 20141101 + LLVM 3.4.2, File-System: ext4, Screen Resolution: 1920x1080 -Ofast -march=native: Processor: Intel Xeon E5-1680 v3 @ 3.80GHz (16 Cores), Motherboard: MSI X99S SLI PLUS (MS-7885) v1.0, Chipset: Intel Xeon E5 v3/Core, Memory: 16384MB, Disk: 80GB INTEL SSDSCKGW08, Graphics: eVGA NVIDIA GeForce GT 740 1024MB, Audio: Intel C610/X99, Monitor: VE228, Network: Intel Connection OS: Fedora 21, Kernel: 3.18.0-0.rc3.git2.2.fc22.1.x86_64 (x86_64), Desktop: GNOME Shell 3.14.1.5, Display Server: X Server 1.16.1, Display Driver: nouveau 1.0.10, OpenGL: 3.3 Mesa 10.3.2 Gallium 0.4, Compiler: GCC 4.9.2 20141101 + LLVM 3.4.2, File-System: ext4, Screen Resolution: 1920x1080 LAME MP3 Encoding 3.99.3 WAV To MP3 Seconds < Lower Is Better -O0 .................. 34.07 |================================================= -O2 .................. 12.91 |=================== -O3 .................. 12.32 |================== -O3 -march=native .... 11.38 |================ -Ofast -march=native . 10.44 |=============== FLAC Audio Encoding 1.3.0 WAV To FLAC Seconds < Lower Is Better -O0 .................. 20.89 |================================================= -O2 .................. 5.58 |============= -O3 .................. 4.77 |=========== -O3 -march=native .... 3.93 |========= -Ofast -march=native . 3.89 |========= Primesieve 5.4 1e12 Prime Number Generation Seconds < Lower Is Better -O0 ..................... 214.14 |============================================= -O2 ..................... 45.86 |========== -O3 ..................... 45.90 |========== -O3 -march=native ....... 43.20 |========= -O3 -march=native -flto . 43.16 |========= -Ofast -march=native .... 43.19 |========= C-Ray 1.1 Total Time Seconds < Lower Is Better -O0 ..................... 36.27 |============================================== -O2 ..................... 23.78 |============================== -O3 ..................... 13.69 |================= -O3 -march=native ....... 9.33 |============ -O3 -march=native -flto . 9.32 |============ -Ofast -march=native .... 8.46 |=========== Timed Apache Compilation 2.4.7 Time To Compile Seconds < Lower Is Better -O0 .................. 17.63 |========================== -O2 .................. 30.34 |============================================= -O3 .................. 32.72 |================================================= -O3 -march=native .... 26.82 |======================================== -Ofast -march=native . 26.39 |======================================== Himeno Benchmark 3.0 Poisson Pressure Solver MFLOPS > Higher Is Better -O0 ..................... 461.42 |========= -O2 ..................... 1996.72 |======================================== -O3 ..................... 1997.60 |======================================== -O3 -march=native ....... 1909.10 |====================================== -O3 -march=native -flto . 2222.87 |============================================ -Ofast -march=native .... 2119.06 |========================================== GraphicsMagick 1.3.19 Operation: Local Adaptive Thresholding Iterations Per Minute > Higher Is Better -O0 .................. 19 |=========== -O2 .................. 92 |==================================================== -O3 .................. 92 |==================================================== -O3 -march=native .... 92 |==================================================== -Ofast -march=native . 92 |==================================================== GraphicsMagick 1.3.19 Operation: HWB Color Space Iterations Per Minute > Higher Is Better -O0 .................. 98 |========================= -O2 .................. 144 |===================================== -O3 .................. 141 |===================================== -O3 -march=native .... 168 |=========================================== -Ofast -march=native . 197 |=================================================== GraphicsMagick 1.3.19 Operation: Resizing Iterations Per Minute > Higher Is Better -O0 .................. 98 |=========================== -O2 .................. 177 |================================================== -O3 .................. 172 |================================================ -O3 -march=native .... 182 |=================================================== -Ofast -march=native . 182 |=================================================== GraphicsMagick 1.3.19 Operation: Sharpen Iterations Per Minute > Higher Is Better -O0 .................. 60 |====================== -O2 .................. 130 |=============================================== -O3 .................. 128 |============================================== -O3 -march=native .... 141 |=================================================== -Ofast -march=native . 141 |=================================================== GraphicsMagick 1.3.19 Operation: Blur Iterations Per Minute > Higher Is Better -O0 .................. 66 |======================== -O2 .................. 100 |===================================== -O3 .................. 104 |====================================== -O3 -march=native .... 138 |=================================================== -Ofast -march=native . 137 |=================================================== Gcrypt Library 1.4.4 CAMELLIA256-ECB Cipher Microseconds < Lower Is Better -O0 ..................... 7317 |=============================================== -O2 ..................... 1897 |============ -O3 ..................... 1972 |============= -O3 -march=native ....... 1792 |============ -O3 -march=native -flto . 1997 |============= SciMark 2.0 Computational Test: Fast Fourier Transform Mflops > Higher Is Better -O0 ..................... 443.48 |========================================== -O2 ..................... 472.28 |============================================= -O3 ..................... 444.88 |========================================== -O3 -march=native ....... 472.98 |============================================= -O3 -march=native -flto . 467.81 |============================================= -Ofast -march=native .... 458.28 |============================================ SciMark 2.0 Computational Test: Monte Carlo Mflops > Higher Is Better -O0 ..................... 599.65 |======================================== -O2 ..................... 600.06 |======================================== -O3 ..................... 591.22 |======================================== -O3 -march=native ....... 589.31 |======================================= -O3 -march=native -flto . 673.31 |============================================= -Ofast -march=native .... 602.01 |========================================