LLVM Clang 3.2 Loop Vectorizer Intel Core i7-3960X testing of the automatic loop vectorizer in LLVM 3.2 with the Clang compiler. Benchmarking by Michael Larabel for a future article on phoronix.com. Loop Vectorization: Processor: Intel Core i7-3960X @ 3.30GHz (12 Cores), Motherboard: Intel DX79SI, Chipset: Intel Xeon E5/Core, Memory: 8192MB, Disk: 64GB OCZ VERTEX, Graphics: AMD Radeon HD 4650 512MB, Audio: Realtek ALC892, Monitor: DELL S2409W, Network: Intel 82579LM Gigabit Connection OS: Ubuntu 12.10, Kernel: 3.5.0-17-generic (x86_64), Desktop: Unity 6.8.0, Display Server: X Server 1.13.0, Display Driver: radeon 6.99.99, OpenGL: 2.1 Mesa 9.0 Gallium 0.4, Compiler: Clang 3.2 (SVN 166775) + LLVM 3.2svn, File-System: ext4, Screen Resolution: 1920x1080 Default: Processor: Intel Core i7-3960X @ 3.30GHz (12 Cores), Motherboard: Intel DX79SI, Chipset: Intel Xeon E5/Core, Memory: 8192MB, Disk: 64GB OCZ VERTEX, Graphics: AMD Radeon HD 4650 512MB, Audio: Realtek ALC892, Monitor: DELL S2409W, Network: Intel 82579LM Gigabit Connection OS: Ubuntu 12.10, Kernel: 3.5.0-17-generic (x86_64), Desktop: Unity 6.8.0, Display Server: X Server 1.13.0, Display Driver: radeon 6.99.99, OpenGL: 2.1 Mesa 9.0 Gallium 0.4, Compiler: Clang 3.2 (SVN 166775) + LLVM 3.2svn, File-System: ext4, Screen Resolution: 1920x1080 PostgreSQL pgbench 8.4.11 TPC-B Transactions Per Second TPS > Higher Is Better Loop Vectorization . 336.65 |================================================== Default ............ 324.77 |================================================ Smallpt 1.0 Global Illumination Renderer; 100 Samples Seconds < Lower Is Better Loop Vectorization . 157 |===================================================== Default ............ 153 |==================================================== C-Ray 1.1 Total Time Seconds < Lower Is Better Loop Vectorization . 23.15 |=================================================== Default ............ 20.92 |============================================== Himeno Benchmark 3.0 Poisson Pressure Solver MFLOPS > Higher Is Better Loop Vectorization . 1554.66 |================================================ Default ............ 1595.35 |================================================= GraphicsMagick 1.3.16 Operation: Local Adaptive Thresholding Iterations Per Minute > Higher Is Better Loop Vectorization . 20 |========================= Default ............ 44 |====================================================== GraphicsMagick 1.3.16 Operation: HWB Color Space Iterations Per Minute > Higher Is Better Loop Vectorization . 122 |===================================================== Default ............ 123 |===================================================== GraphicsMagick 1.3.16 Operation: Resizing Iterations Per Minute > Higher Is Better Loop Vectorization . 84 |=================================================== Default ............ 89 |====================================================== GraphicsMagick 1.3.16 Operation: Sharpen Iterations Per Minute > Higher Is Better Loop Vectorization . 31 |====================================================== Default ............ 31 |====================================================== GraphicsMagick 1.3.16 Operation: Blur Iterations Per Minute > Higher Is Better Loop Vectorization . 73 |================================================ Default ............ 82 |====================================================== Timed HMMer Search 2.3.2 Pfam Database Search Seconds < Lower Is Better Loop Vectorization . 15.37 |=================================================== Default ............ 15.35 |===================================================