AMD Bulldozer Compiler Tests AMD FX-4100 Bulldozer Quad-core compiler benchmarking under Ubuntu Linux with GCC 4.5 and GCC 4.6. Testing for a future article on Phoronix.com by Michael Larabel. Thanks to Daniel Newkirk for the SSH access to this AMD Bulldozer system due to lack of AMD FX CPUs at Phoronix. Bulldozer GCC 4.5.2: Processor: AMD FX-4100 @ 3.60GHz (4 Cores), Motherboard: ASUS M5A97 EVO, Chipset: ATI RD890 PCI to PCI bridge, Memory: 16384MB, Disk: 60GB SSD G2 64 + 2000GB SAMSUNG HD204UI + 1000GB Western Digital WD1001FALS-0, Graphics: NVIDIA GeForce 8400 GS, Audio: Realtek ALC892, Network: Realtek RTL8111/8168B OS: Ubuntu 11.04, Kernel: 2.6.38-11-generic (x86_64), Display Driver: NVIDIA 1.0.0, Compiler: GCC 4.5.2, File-System: xfs Bulldozer GCC 4.6.1: Processor: AMD FX-4100 @ 3.60GHz (4 Cores), Motherboard: ASUS M5A97 EVO, Chipset: ATI RD890 PCI to PCI bridge, Memory: 16384MB, Disk: 60GB SSD G2 64 + 2000GB SAMSUNG HD204UI + 1000GB Western Digital WD1001FALS-0, Graphics: NVIDIA GeForce 8400 GS, Audio: Realtek ALC892, Network: Realtek RTL8111/8168B OS: Ubuntu 11.04, Kernel: 2.6.38-11-generic (x86_64), Display Driver: NVIDIA 1.0.0, Compiler: GCC 4.6.1, File-System: xfs phii840: Processor: AMD Phenom II X4 840 @ 3.21GHz (4 Cores), Motherboard: Gigabyte GA-MA790GP-DS4H, Chipset: AMD RS780 + SB700/SB800, Memory: 4096MB, Disk: 1000GB SAMSUNG HD103SJ + 500GB Western Digital WDC WD5000AAKS-0 + 1000GB SAMSUNG HD103UJ, Graphics: NVIDIA Quadro FX 570 256MB (460/400MHz), Audio: Realtek ALC889A, Network: Realtek RTL8111/8168B OS: Fedora 14, Kernel: 2.6.35.10-74.fc14.x86_64 (x86_64), Desktop: GNOME 2.32.0, Display Server: X Server 1.9.5, Display Driver: NVIDIA 270.41.06, OpenGL: 3.3.0 NVIDIA 270.41.06, File-System: ext4, Screen Resolution: 1920x1080 C-Ray 1.1 Total Time Seconds < Lower Is Better Bulldozer GCC 4.5.2 . 106.75 |================================================= Bulldozer GCC 4.6.1 . 105.49 |================================================ phii840 ............. 107.42 |================================================= Smallpt 1.0 Global Illumination Renderer; 100 Samples Seconds < Lower Is Better Bulldozer GCC 4.5.2 . 211 |=================================================== Bulldozer GCC 4.6.1 . 215 |==================================================== phii840 ............. 194 |=============================================== POV-Ray 3.6.1 Total Time Seconds < Lower Is Better Bulldozer GCC 4.5.2 . 913 |=========================================== Bulldozer GCC 4.6.1 . 916 |============================================ phii840 ............. 1071 |=================================================== TSCP 1.81 AI Chess Performance Nodes Per Second > Higher Is Better Bulldozer GCC 4.5.2 . 293780 |================================================= Bulldozer GCC 4.6.1 . 286148 |================================================ phii840 ............. 270068 |============================================= GraphicsMagick 1.3.12 Operation: HWB Color Space Iterations Per Minute > Higher Is Better Bulldozer GCC 4.5.2 . 123 |================================================ Bulldozer GCC 4.6.1 . 134 |==================================================== phii840 ............. 122 |=============================================== GraphicsMagick 1.3.12 Operation: Blur Iterations Per Minute > Higher Is Better Bulldozer GCC 4.5.2 . 87 |=================================================== Bulldozer GCC 4.6.1 . 90 |===================================================== phii840 ............. 63 |===================================== GraphicsMagick 1.3.12 Operation: Local Adaptive Thresholding Iterations Per Minute > Higher Is Better Bulldozer GCC 4.5.2 . 44 |==================================================== Bulldozer GCC 4.6.1 . 43 |=================================================== phii840 ............. 45 |===================================================== GraphicsMagick 1.3.12 Operation: Resizing Iterations Per Minute > Higher Is Better Bulldozer GCC 4.5.2 . 92 |============================================== Bulldozer GCC 4.6.1 . 96 |================================================ phii840 ............. 103 |==================================================== GraphicsMagick 1.3.12 Operation: Sharpen Iterations Per Minute > Higher Is Better Bulldozer GCC 4.5.2 . 44 |========================================== Bulldozer GCC 4.6.1 . 55 |===================================================== phii840 ............. 50 |================================================ John The Ripper 1.7.8 Test: Traditional DES Real C/S > Higher Is Better Bulldozer GCC 4.5.2 . 3523667 |================================================ Bulldozer GCC 4.6.1 . 3321333 |============================================= phii840 ............. 2787000 |====================================== John The Ripper 1.7.8 Test: MD5 Real C/S > Higher Is Better Bulldozer GCC 4.5.2 . 14082 |================================================ Bulldozer GCC 4.6.1 . 13598 |============================================== phii840 ............. 14677 |================================================== John The Ripper 1.7.8 Test: Blowfish Real C/S > Higher Is Better Bulldozer GCC 4.5.2 . 866 |================================================= Bulldozer GCC 4.6.1 . 895 |=================================================== phii840 ............. 910 |==================================================== OpenSSL 1.0.0a RSA 4096-bit Performance Signs Per Second > Higher Is Better Bulldozer GCC 4.5.2 . 63.65 |================================================== Bulldozer GCC 4.6.1 . 63.83 |================================================== phii840 ............. 60.85 |================================================ Gcrypt Library 1.4.4 CAMELLIA256-ECB Cipher Microseconds < Lower Is Better Bulldozer GCC 4.5.2 . 3777 |=================================================== Bulldozer GCC 4.6.1 . 2503 |================================== 7-Zip Compression 9.13 Compress Speed Test MIPS > Higher Is Better Bulldozer GCC 4.5.2 . 8877 |=================================================== Bulldozer GCC 4.6.1 . 8783 |================================================== phii840 ............. 8885 |=================================================== LAME MP3 Encoding 3.98.2 WAV To MP3 Seconds < Lower Is Better Bulldozer GCC 4.5.2 . 28.69 |================================================ Bulldozer GCC 4.6.1 . 29.76 |================================================== phii840 ............. 29.85 |================================================== FLAC Audio Encoding 1.2.1 WAV To FLAC Seconds < Lower Is Better Bulldozer GCC 4.5.2 . 8.75 |=================================================== Bulldozer GCC 4.6.1 . 8.66 |================================================== phii840 ............. 8.51 |================================================== FFmpeg 0.8.2 AVI To NTSC VCD Seconds < Lower Is Better Bulldozer GCC 4.5.2 . 16.50 |================================================ Bulldozer GCC 4.6.1 . 17.06 |================================================== phii840 ............. 13.64 |======================================== x264 2010-11-22 H.264 Video Encoding Frames Per Second > Higher Is Better Bulldozer GCC 4.5.2 . 50.19 |================================================ Bulldozer GCC 4.6.1 . 50.89 |================================================ phii840 ............. 52.57 |================================================== Timed MAFFT Alignment 6.706 Multiple Sequence Alignment Seconds < Lower Is Better Bulldozer GCC 4.5.2 . 28.86 |============================================= Bulldozer GCC 4.6.1 . 28.83 |============================================= phii840 ............. 32.02 |================================================== NAS Parallel Benchmarks 3.3 Test / Class: BT.A Total Mop/s > Higher Is Better Bulldozer GCC 4.5.2 . 5982.31 |============================================ phii840 ............. 6488.78 |================================================ NAS Parallel Benchmarks 3.3 Test / Class: CG.B Total Mop/s > Higher Is Better Bulldozer GCC 4.5.2 . 1788.94 |================================================ phii840 ............. 532.17 |============== NAS Parallel Benchmarks 3.3 Test / Class: EP.B Total Mop/s > Higher Is Better Bulldozer GCC 4.5.2 . 78.97 |============================================ phii840 ............. 88.81 |================================================== NAS Parallel Benchmarks 3.3 Test / Class: FT.B Total Mop/s > Higher Is Better Bulldozer GCC 4.5.2 . 3785.19 |================================================ phii840 ............. 2754.65 |=================================== NAS Parallel Benchmarks 3.3 Test / Class: IS.C Total Mop/s > Higher Is Better Bulldozer GCC 4.5.2 . 99.91 |================================================== Bulldozer GCC 4.6.1 . 99.50 |================================================== phii840 ............. 79.57 |======================================== NAS Parallel Benchmarks 3.3 Test / Class: LU.A Total Mop/s > Higher Is Better Bulldozer GCC 4.5.2 . 6058.10 |================================================ phii840 ............. 2601.13 |===================== NAS Parallel Benchmarks 3.3 Test / Class: MG.B Total Mop/s > Higher Is Better Bulldozer GCC 4.5.2 . 3645.03 |================================================ phii840 ............. 2794.51 |===================================== NAS Parallel Benchmarks 3.3 Test / Class: SP.A Total Mop/s > Higher Is Better Bulldozer GCC 4.5.2 . 3735.79 |================================================ phii840 ............. 2171.99 |============================ NAS Parallel Benchmarks 3.3 Test / Class: UA.A Total Mop/s > Higher Is Better Bulldozer GCC 4.5.2 . 21.48 |================================================== phii840 ............. 17.61 |========================================= CLOMP 3.3 Static OMP Speedup Speedup > Higher Is Better Bulldozer GCC 4.5.2 . 2.31 |========================================== Bulldozer GCC 4.6.1 . 2.31 |========================================== phii840 ............. 2.78 |=================================================== SciMark 2.0 Computational Test: Composite Mflops > Higher Is Better Bulldozer GCC 4.5.2 . 738.80 |================================================= Bulldozer GCC 4.6.1 . 735.23 |================================================= phii840 ............. 475.59 |================================ SciMark 2.0 Computational Test: Fast Fourier Transform Mflops > Higher Is Better Bulldozer GCC 4.5.2 . 72.89 |================================================== Bulldozer GCC 4.6.1 . 69.17 |=============================================== phii840 ............. 66.13 |============================================= SciMark 2.0 Computational Test: Jacobi Successive Over-Relaxation Mflops > Higher Is Better Bulldozer GCC 4.5.2 . 690.52 |================================================ Bulldozer GCC 4.6.1 . 690.51 |================================================ phii840 ............. 704.81 |================================================= SciMark 2.0 Computational Test: Monte Carlo Mflops > Higher Is Better Bulldozer GCC 4.5.2 . 345.55 |================================================= Bulldozer GCC 4.6.1 . 336.07 |================================================ phii840 ............. 337.66 |================================================ SciMark 2.0 Computational Test: Sparse Matrix Multiply Mflops > Higher Is Better Bulldozer GCC 4.5.2 . 968.47 |================================================= Bulldozer GCC 4.6.1 . 940.56 |================================================ phii840 ............. 571.28 |============================= SciMark 2.0 Computational Test: Dense LU Matrix Factorization Mflops > Higher Is Better Bulldozer GCC 4.5.2 . 1616.56 |=============================================== Bulldozer GCC 4.6.1 . 1639.84 |================================================ phii840 ............. 698.09 |==================== Fhourstones 3.1 Complex Connect-4 Solving Kpos / sec > Higher Is Better Bulldozer GCC 4.5.2 . 9425.30 |================================================ Bulldozer GCC 4.6.1 . 9221.03 |=============================================== phii840 ............. 8830.77 |============================================= N-Queens 1.0 Elapsed Time Seconds < Lower Is Better Bulldozer GCC 4.5.2 . 281.74 |=========================================== Bulldozer GCC 4.6.1 . 322.56 |================================================= phii840 ............. 241.08 |=====================================