Tests for a future article.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 1801075-AL-RETPOLINE03 retpoline-testing - Phoronix Test Suite retpoline-testing Tests for a future article.
HTML result view exported from: https://openbenchmarking.org/result/1801075-AL-RETPOLINE03&export=txt&sor&grs .
Processor Motherboard Chipset Memory Disk Graphics Monitor Network OS Kernel Desktop Display Driver OpenCL Compiler File-System Screen Resolution EPYC 7601 2 x Xeon Gold 6138 noretpoline Retpoline Retpoline + GCC no retpoline Retpoline Retpoline + GCC AMD EPYC 7601 32-Core @ 2.20GHz (32 Cores / 64 Threads) TYAN B8026T70AE24HR AMD Device 1450 129024MB 280GB INTEL SSDPE21D280GA ASPEED ASPEED Family VE228 Broadcom Limited NetXtreme BCM5720 Gigabit PCIe Ubuntu 17.10 4.14.0-phx-retpoline (x86_64) GNOME Shell 3.26.1 modesetting 1.19.5 OpenCL 1.2 pocl 1.0 LLVM 5.0.0 GCC 7.2.0 + Clang 5.0.0-3 + LLVM 5.0.0 ext4 1920x1080 4.14.0-phx-retpoline-gcc-retpo (x86_64) 2 x Intel Xeon Gold 6138 @ 3.70GHz (40 Cores / 80 Threads) TYAN S7106 Intel Device 2020 96256MB 256GB Samsung SSD 850 + 2000GB Seagate ST2000DM006-2DM1 + 2 x 120GB TOSHIBA-TR150 Intel I210 Gigabit Connection 4.14.0-phx-retpoline (x86_64) GCC 7.2.0 4.14.0-phx-retpoline-gcc-retpo (x86_64) OpenBenchmarking.org Compiler Details - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v Disk Details - EPYC 7601: noretpoline: NONE / data=ordered,errors=remount-ro,relatime,rw - EPYC 7601: Retpoline: NONE / data=ordered,errors=remount-ro,relatime,rw - EPYC 7601: Retpoline + GCC: NONE / data=ordered,errors=remount-ro,relatime,rw - 2 x Xeon Gold 6138: no retpoline: CFQ / data=ordered,errors=remount-ro,relatime,rw - 2 x Xeon Gold 6138: Retpoline: CFQ / data=ordered,errors=remount-ro,relatime,rw - 2 x Xeon Gold 6138: Retpoline + GCC: CFQ / data=ordered,errors=remount-ro,relatime,rw Processor Details - EPYC 7601: noretpoline: Scaling Governor: acpi-cpufreq ondemand - EPYC 7601: Retpoline: Scaling Governor: acpi-cpufreq ondemand - EPYC 7601: Retpoline + GCC: Scaling Governor: acpi-cpufreq ondemand - 2 x Xeon Gold 6138: no retpoline: Scaling Governor: intel_pstate powersave - 2 x Xeon Gold 6138: Retpoline: Scaling Governor: intel_pstate powersave - 2 x Xeon Gold 6138: Retpoline + GCC: Scaling Governor: intel_pstate powersave System Details - Python 2.7.14.
fs-mark: 4000 Files, 32 Sub Dirs, 1MB Size scikit-learn: fio: Seq Read - Libaio - No - Yes - 2MB - Default Test Directory fs-mark: 1000 Files, 1MB Size fio: Rand Read - Libaio - No - Yes - 2MB - Default Test Directory fio: Rand Write - Libaio - No - Yes - 2MB - Default Test Directory fio: Seq Write - Libaio - No - Yes - 2MB - Default Test Directory fio: Rand Write - Libaio - No - Yes - 4KB - Default Test Directory fio: Rand Read - Libaio - No - Yes - 4KB - Default Test Directory fio: Seq Write - Libaio - No - Yes - 4KB - Default Test Directory fio: Seq Read - Libaio - No - Yes - 4KB - Default Test Directory glibc-bench: sqrt pgbench: Buffer Test - Normal Load - Read Write t-test1: 1 t-test1: 2 glibc-bench: ffs parboil: OpenMP MRI Gridding john-the-ripper: Blowfish redis: SET pybench: Total For Average Test Times cachebench: Read apache: Static Web Page Serving lzbench: XZ 0 - Compression compilebench: Initial Create stockfish: Total Time glibc-bench: pthread_once build-apache: Time To Compile compress-lzma: 256MB File Compression parboil: OpenMP CUTCP cachebench: Write ebizzy: rodinia: OpenMP CFD Solver compilebench: Compile cachebench: Read / Modify / Write rodinia: OpenMP LavaMD c-ray: Total Time lzbench: Zstd 1 - Compression tjbench: Decompression Throughput pgbench: Buffer Test - Normal Load - Read Only redis: GET redis: LPUSH redis: SADD redis: LPOP build-linux-kernel: Time To Compile EPYC 7601 2 x Xeon Gold 6138 noretpoline Retpoline Retpoline + GCC no retpoline Retpoline Retpoline + GCC 581.87 34.00 2511.77 597.93 2509.83 1972.70 1971.10 1059.73 1176.20 1081.23 1183.53 4.70 37.76 14.38 4.83 283.16 35458 1131889.69 1794 2205.73 16587.34 24 410.85 4507 4.83 31.71 329.57 2.71 21622.77 1093230 10.97 1696.87 22924.24 31.65 3.50 335 140.83 1394712.79 1117859.50 1233207.56 1520530.25 38.30 629.27 33.91 2618.67 639.77 2615.60 2157.60 2158.03 1046.93 1165.97 1091.50 1160.83 4.70 36.68 14.46 4.83 283.25 35714 1161527.67 1801 2215.24 16838.76 24 406.41 4507 4.83 32.21 328.95 2.73 21556.98 1073587 10.76 1691.37 22918.96 31.86 3.45 336 140.99 1362255.92 1147686.87 1293172.50 1394465.79 38.44 631.30 34.47 2618.20 619.67 2637.27 2158.27 2156.87 1048.77 1128.57 1072.57 1193.80 4.70 38.18 14.66 4.83 288.36 35715 1162399.91 1795 2214.25 16716.48 24 406.91 4501 4.83 32.12 329.00 2.70 21835.76 952691 11.08 1693.10 22900.81 31.87 3.48 335 140.87 1352327.02 1123252.42 1249051.08 1230732.38 38.52 111.02 184.75 519.38 128.67 543.05 523.89 514.75 338.58 383.14 415.49 457.26 12.30 2136.64 65.91 23.12 3.26 407.91 49652 1564339.04 1307 2973.08 22258.72 32 533.70 3437 3.87 26.24 281.84 2.35 24954.96 946321 9.73 1496.78 25777.79 28.50 3.16 361 145.81 599258.14 1627717.71 1357638.81 1589703.69 1423431.25 30.25 136.67 185.68 518.49 133.73 543.18 514.67 526.12 341.01 384.79 413.25 454.44 9.95 2467.38 66.56 23.29 3.26 413.85 48399 1525833.21 1306 2872.09 21347.64 30 508.92 3654 3.80 26.31 281.25 2.38 23882.88 954773 9.93 1694.04 25061.33 28.70 3.15 369 154.02 597483.03 1738340.06 1382896.04 1573307.83 1486377.46 30.46 111.17 186.08 518.48 129.63 543.37 512.18 527.34 341.98 388.19 416.94 456.92 9.96 3976.32 64.45 23.04 3.26 405.25 51089 1399642.27 1315 2839.79 18757.65 31 478.76 3613 4.49 26.32 282.04 2.38 23753.79 950671 10.54 1637.02 24928.44 28.43 3.18 370 146.51 577641.99 1574731.42 1454926.31 1604656.60 1347143.44 30.63 OpenBenchmarking.org
FS-Mark Test: 4000 Files, 32 Sub Dirs, 1MB Size EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org Files/s, More Is Better FS-Mark 3.3 Test: 4000 Files, 32 Sub Dirs, 1MB Size Retpoline + GCC Retpoline noretpoline no retpoline 140 280 420 560 700 SE +/- 4.00, N = 3 SE +/- 1.74, N = 3 SE +/- 1.13, N = 3 SE +/- 0.54, N = 3 SE +/- 7.23, N = 6 SE +/- 2.97, N = 6 631.30 629.27 581.87 136.67 111.17 111.02 1. (CC) gcc options: -static
Scikit-Learn EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org Seconds, Fewer Is Better Scikit-Learn 0.17.1 Retpoline noretpoline Retpoline + GCC no retpoline 40 80 120 160 200 SE +/- 0.02, N = 3 SE +/- 0.06, N = 3 SE +/- 0.48, N = 3 SE +/- 0.99, N = 3 SE +/- 0.29, N = 3 SE +/- 2.58, N = 3 33.91 34.00 34.47 184.75 185.68 186.08
Flexible IO Tester Type: Sequential Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 2.1.13 Type: Sequential Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Retpoline + GCC Retpoline noretpoline no retpoline 300 600 900 1200 1500 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 1306 1306 1252 250 250 250 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 2.1.13 Type: Sequential Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Retpoline Retpoline + GCC noretpoline no retpoline 600 1200 1800 2400 3000 SE +/- 0.03, N = 3 SE +/- 0.10, N = 3 SE +/- 0.03, N = 3 SE +/- 0.72, N = 3 SE +/- 0.25, N = 3 SE +/- 0.05, N = 3 2618.67 2618.20 2511.77 519.38 518.49 518.48 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Random Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 2.1.13 Type: Random Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Retpoline + GCC Retpoline noretpoline no retpoline 300 600 900 1200 1500 SE +/- 11.00, N = 3 SE +/- 0.33, N = 3 SE +/- 0.67, N = 3 1315 1304 1251 262 262 262 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
FS-Mark Test: 1000 Files, 1MB Size EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org Files/s, More Is Better FS-Mark 3.3 Test: 1000 Files, 1MB Size Retpoline Retpoline + GCC noretpoline no retpoline 140 280 420 560 700 SE +/- 1.90, N = 3 SE +/- 3.59, N = 3 SE +/- 5.75, N = 3 SE +/- 0.20, N = 3 SE +/- 2.15, N = 6 SE +/- 2.17, N = 3 639.77 619.67 597.93 133.73 129.63 128.67 1. (CC) gcc options: -static
Flexible IO Tester Type: Random Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 2.1.13 Type: Random Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Retpoline + GCC Retpoline noretpoline no retpoline 600 1200 1800 2400 3000 SE +/- 22.27, N = 3 SE +/- 0.45, N = 3 SE +/- 0.24, N = 3 SE +/- 0.10, N = 3 SE +/- 0.14, N = 3 SE +/- 1.17, N = 3 2637.27 2615.60 2509.83 543.37 543.18 543.05 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Random Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 2.1.13 Type: Random Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Retpoline + GCC Retpoline noretpoline no retpoline 200 400 600 800 1000 SE +/- 1.67, N = 3 SE +/- 1.53, N = 3 SE +/- 0.67, N = 3 SE +/- 0.58, N = 3 SE +/- 1.20, N = 3 SE +/- 1.20, N = 3 1076 1075 983 252 248 246 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 2.1.13 Type: Sequential Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Retpoline + GCC Retpoline noretpoline no retpoline 200 400 600 800 1000 SE +/- 1.86, N = 3 SE +/- 1.86, N = 3 SE +/- 0.58, N = 3 SE +/- 4.34, N = 4 1075 1075 982 254 253 248 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Random Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 2.1.13 Type: Random Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Retpoline + GCC Retpoline noretpoline no retpoline 500 1000 1500 2000 2500 SE +/- 3.18, N = 3 SE +/- 3.16, N = 3 SE +/- 1.43, N = 3 SE +/- 0.94, N = 3 SE +/- 2.04, N = 3 SE +/- 2.16, N = 3 2158.27 2157.60 1972.70 523.89 514.67 512.18 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 2.1.13 Type: Sequential Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Retpoline Retpoline + GCC noretpoline no retpoline 500 1000 1500 2000 2500 SE +/- 3.77, N = 3 SE +/- 4.03, N = 3 SE +/- 1.26, N = 3 SE +/- 0.39, N = 3 SE +/- 0.90, N = 3 SE +/- 8.78, N = 4 2158.03 2156.87 1971.10 527.34 526.12 514.75 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Random Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 2.1.13 Type: Random Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory noretpoline Retpoline + GCC Retpoline no retpoline 60K 120K 180K 240K 300K SE +/- 2932.09, N = 3 SE +/- 2976.09, N = 3 SE +/- 4419.58, N = 6 SE +/- 451.27, N = 3 SE +/- 444.91, N = 3 SE +/- 484.83, N = 3 271273 268542 268041 85492 85249 84641 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Random Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 2.1.13 Type: Random Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory noretpoline Retpoline Retpoline + GCC no retpoline 60K 120K 180K 240K 300K SE +/- 4402.78, N = 3 SE +/- 3375.56, N = 3 SE +/- 591.14, N = 3 SE +/- 504.90, N = 3 SE +/- 309.68, N = 3 SE +/- 1071.43, N = 3 301037 298428 288951 97043 96194 95782 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Random Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 2.1.13 Type: Random Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory noretpoline Retpoline + GCC Retpoline no retpoline 200 400 600 800 1000 SE +/- 11.43, N = 3 SE +/- 11.31, N = 3 SE +/- 17.28, N = 6 SE +/- 1.80, N = 3 SE +/- 1.78, N = 3 SE +/- 1.94, N = 3 1059.73 1048.77 1046.93 341.98 341.01 338.58 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Random Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 2.1.13 Type: Random Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory noretpoline Retpoline Retpoline + GCC no retpoline 300 600 900 1200 1500 SE +/- 17.24, N = 3 SE +/- 13.29, N = 3 SE +/- 1.94, N = 3 SE +/- 2.02, N = 3 SE +/- 1.24, N = 3 SE +/- 4.29, N = 3 1176.20 1165.97 1128.57 388.19 384.79 383.14 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 2.1.13 Type: Sequential Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory Retpoline noretpoline Retpoline + GCC no retpoline 60K 120K 180K 240K 300K SE +/- 4389.73, N = 6 SE +/- 3663.70, N = 3 SE +/- 1888.66, N = 3 SE +/- 361.30, N = 3 SE +/- 614.11, N = 3 SE +/- 241.21, N = 3 279404 276778 274530 104232 103868 103308 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 2.1.13 Type: Sequential Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory Retpoline + GCC noretpoline Retpoline no retpoline 70K 140K 210K 280K 350K SE +/- 4452.90, N = 3 SE +/- 4646.93, N = 3 SE +/- 1598.66, N = 3 SE +/- 246.03, N = 3 SE +/- 379.04, N = 3 SE +/- 33.79, N = 3 305665 302950 297144 114311 114227 113608 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 2.1.13 Type: Sequential Write - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory Retpoline noretpoline Retpoline + GCC no retpoline 200 400 600 800 1000 SE +/- 17.15, N = 6 SE +/- 14.34, N = 3 SE +/- 7.38, N = 3 SE +/- 1.44, N = 3 SE +/- 2.46, N = 3 SE +/- 0.97, N = 3 1091.50 1081.23 1072.57 416.94 415.49 413.25 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 2.1.13 Type: Sequential Read - IO Engine: Libaio - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test Directory Retpoline + GCC noretpoline Retpoline no retpoline 300 600 900 1200 1500 SE +/- 17.58, N = 3 SE +/- 18.11, N = 3 SE +/- 6.24, N = 3 SE +/- 0.98, N = 3 SE +/- 1.52, N = 3 SE +/- 0.14, N = 3 1193.80 1183.53 1160.83 457.26 456.92 454.44 1. (CC) gcc options: -rdynamic -std=gnu99 -O3 -ffast-math -include -lrt -laio -lz -lm -lpthread -ldl
glibc bench Benchmark: sqrt EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org nanoseconds, Fewer Is Better glibc bench 1.0 Benchmark: sqrt noretpoline Retpoline Retpoline + GCC no retpoline 3 6 9 12 15 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.62, N = 6 SE +/- 0.62, N = 6 SE +/- 0.00, N = 3 4.70 4.70 4.70 9.95 9.96 12.30
PostgreSQL pgbench Scaling: Buffer Test - Test: Normal Load - Mode: Read Write OpenBenchmarking.org TPS, More Is Better PostgreSQL pgbench 10.0 Scaling: Buffer Test - Test: Normal Load - Mode: Read Write Retpoline + GCC Retpoline no retpoline 900 1800 2700 3600 4500 SE +/- 15.09, N = 3 SE +/- 49.01, N = 6 SE +/- 40.05, N = 3 3976.32 2467.38 2136.64 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -fPIC -lpgcommon -lpgport -lpthread -lrt -lcrypt -ldl -lm
t-test1 Threads: 1 EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org Seconds, Fewer Is Better t-test1 2017-01-13 Threads: 1 Retpoline noretpoline Retpoline + GCC no retpoline 15 30 45 60 75 SE +/- 0.10, N = 3 SE +/- 0.09, N = 3 SE +/- 0.12, N = 3 SE +/- 0.14, N = 3 SE +/- 0.15, N = 3 SE +/- 0.09, N = 3 36.68 37.76 38.18 64.45 65.91 66.56 1. (CC) gcc options: -pthread
t-test1 Threads: 2 EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org Seconds, Fewer Is Better t-test1 2017-01-13 Threads: 2 noretpoline Retpoline Retpoline + GCC no retpoline 6 12 18 24 30 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.06, N = 3 SE +/- 0.09, N = 3 SE +/- 0.13, N = 3 SE +/- 0.06, N = 3 14.38 14.46 14.66 23.04 23.12 23.29 1. (CC) gcc options: -pthread
glibc bench Benchmark: ffs 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org nanoseconds, Fewer Is Better glibc bench 1.0 Benchmark: ffs no retpoline Retpoline Retpoline + GCC noretpoline 1.0868 2.1736 3.2604 4.3472 5.434 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 3.26 3.26 3.26 4.83 4.83 4.83
Parboil Test: OpenMP MRI Gridding EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org Seconds, Fewer Is Better Parboil 2.5 Test: OpenMP MRI Gridding noretpoline Retpoline Retpoline + GCC no retpoline 90 180 270 360 450 SE +/- 0.38, N = 3 SE +/- 0.14, N = 3 SE +/- 1.39, N = 3 SE +/- 6.86, N = 4 SE +/- 6.83, N = 3 SE +/- 6.92, N = 3 283.16 283.25 288.36 405.25 407.91 413.85 1. (CXX) g++ options: -lm -lpthread -lgomp -ffast-math -fopenmp
John The Ripper Test: Blowfish 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Real C/S, More Is Better John The Ripper 1.8.0 Test: Blowfish Retpoline + GCC no retpoline Retpoline noretpoline 11K 22K 33K 44K 55K SE +/- 357.46, N = 3 SE +/- 562.61, N = 3 SE +/- 1882.95, N = 6 SE +/- 232.00, N = 3 SE +/- 192.26, N = 3 SE +/- 400.35, N = 3 51089 49652 48399 35715 35714 35458 1. (CC) gcc options: -fopenmp -lcrypt
Redis Test: SET 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Requests Per Second, More Is Better Redis 3.0.1 Test: SET no retpoline Retpoline Retpoline + GCC noretpoline 300K 600K 900K 1200K 1500K SE +/- 28059.92, N = 3 SE +/- 90449.67, N = 6 SE +/- 55833.41, N = 6 SE +/- 5879.00, N = 3 SE +/- 7115.17, N = 3 SE +/- 26854.76, N = 6 1564339.04 1525833.21 1399642.27 1162399.91 1161527.67 1131889.69 1. (CC) gcc options: -ggdb -rdynamic -lm -pthread
PyBench Total For Average Test Times 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Milliseconds, Fewer Is Better PyBench 2008-08-14 Total For Average Test Times Retpoline no retpoline Retpoline + GCC noretpoline 400 800 1200 1600 2000 SE +/- 1.53, N = 3 SE +/- 1.76, N = 3 SE +/- 7.97, N = 3 SE +/- 11.33, N = 3 1306 1307 1315 1794 1795 1801
CacheBench Test: Read 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org MB/s, More Is Better CacheBench Test: Read no retpoline Retpoline Retpoline + GCC noretpoline 600 1200 1800 2400 3000 SE +/- 7.30, N = 3 SE +/- 12.31, N = 3 SE +/- 11.39, N = 3 SE +/- 0.02, N = 3 SE +/- 0.98, N = 3 SE +/- 9.38, N = 3 2973.08 2872.09 2839.79 2215.24 2214.25 2205.73 1. (CC) gcc options: -lrt
Apache Benchmark Static Web Page Serving 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Requests Per Second, More Is Better Apache Benchmark 2.4.7 Static Web Page Serving no retpoline Retpoline Retpoline + GCC noretpoline 5K 10K 15K 20K 25K SE +/- 139.49, N = 3 SE +/- 63.48, N = 3 SE +/- 51.11, N = 3 SE +/- 97.15, N = 3 SE +/- 296.19, N = 3 SE +/- 41.45, N = 3 22258.72 21347.64 18757.65 16838.76 16716.48 16587.34 1. (CC) gcc options: -shared -fPIC -O2 -pthread
lzbench Test: XZ 0 - Process: Compression 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org MB/s, More Is Better lzbench 2017-08-08 Test: XZ 0 - Process: Compression no retpoline Retpoline + GCC Retpoline noretpoline 7 14 21 28 35 SE +/- 0.72, N = 6 SE +/- 0.50, N = 6 SE +/- 0.56, N = 6 32 31 30 24 24 24 1. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3
Compile Bench Test: Initial Create 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org MB/s, More Is Better Compile Bench 0.6 Test: Initial Create no retpoline Retpoline Retpoline + GCC noretpoline 120 240 360 480 600 SE +/- 1.88, N = 3 SE +/- 25.41, N = 3 SE +/- 27.64, N = 3 SE +/- 2.54, N = 3 SE +/- 4.20, N = 3 SE +/- 3.88, N = 3 533.70 508.92 478.76 410.85 406.91 406.41
Stockfish Total Time 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org ms, Fewer Is Better Stockfish 2014-11-26 Total Time no retpoline Retpoline + GCC Retpoline noretpoline 1000 2000 3000 4000 5000 SE +/- 37.22, N = 3 SE +/- 210.25, N = 6 SE +/- 202.73, N = 6 SE +/- 2.65, N = 3 SE +/- 3.48, N = 3 3437 3613 3654 4501 4507 4507 1. (CXX) g++ options: -lpthread -fno-exceptions -fno-rtti -ansi -pedantic -O3 -msse -msse3 -mpopcnt -flto
glibc bench Benchmark: pthread_once 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org nanoseconds, Fewer Is Better glibc bench 1.0 Benchmark: pthread_once Retpoline no retpoline Retpoline + GCC noretpoline 1.0868 2.1736 3.2604 4.3472 5.434 SE +/- 0.00, N = 3 SE +/- 0.06, N = 3 SE +/- 0.24, N = 6 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 3.80 3.87 4.49 4.83 4.83 4.83
Timed Apache Compilation Time To Compile 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Seconds, Fewer Is Better Timed Apache Compilation 2.4.7 Time To Compile no retpoline Retpoline Retpoline + GCC noretpoline 7 14 21 28 35 SE +/- 0.06, N = 3 SE +/- 0.14, N = 3 SE +/- 0.10, N = 3 SE +/- 0.14, N = 3 SE +/- 0.07, N = 3 SE +/- 0.07, N = 3 26.24 26.31 26.32 31.71 32.12 32.21
LZMA Compression 256MB File Compression 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Seconds, Fewer Is Better LZMA Compression 256MB File Compression Retpoline no retpoline Retpoline + GCC noretpoline 70 140 210 280 350 SE +/- 0.60, N = 3 SE +/- 0.34, N = 3 SE +/- 2.80, N = 3 SE +/- 0.22, N = 3 SE +/- 0.42, N = 3 SE +/- 0.18, N = 3 281.25 281.84 282.04 328.95 329.00 329.57 1. (CXX) g++ options: -O2
lzbench Test: XZ 0 - Process: Decompression 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org MB/s, More Is Better lzbench 2017-08-08 Test: XZ 0 - Process: Decompression no retpoline Retpoline Retpoline + GCC noretpoline 20 40 60 80 100 SE +/- 1.03, N = 6 SE +/- 2.00, N = 6 SE +/- 1.77, N = 6 89 87 86 77 77 76 1. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3
Parboil Test: OpenMP CUTCP 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Seconds, Fewer Is Better Parboil 2.5 Test: OpenMP CUTCP no retpoline Retpoline Retpoline + GCC noretpoline 0.6143 1.2286 1.8429 2.4572 3.0715 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 6 2.35 2.38 2.38 2.70 2.71 2.73 1. (CXX) g++ options: -lm -lpthread -lgomp -ffast-math -fopenmp
CacheBench Test: Write 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org MB/s, More Is Better CacheBench Test: Write no retpoline Retpoline Retpoline + GCC noretpoline 5K 10K 15K 20K 25K SE +/- 34.79, N = 3 SE +/- 197.71, N = 3 SE +/- 82.06, N = 3 SE +/- 25.70, N = 3 SE +/- 53.65, N = 3 SE +/- 133.19, N = 3 24954.96 23882.88 23753.79 21835.76 21622.77 21556.98 1. (CC) gcc options: -lrt
ebizzy EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org Records/s, More Is Better ebizzy 0.3 noretpoline Retpoline Retpoline + GCC no retpoline 200K 400K 600K 800K 1000K SE +/- 19950.45, N = 3 SE +/- 26430.87, N = 6 SE +/- 23463.01, N = 6 SE +/- 13516.78, N = 6 SE +/- 13692.48, N = 5 SE +/- 7835.23, N = 3 1093230 1073587 952691 954773 950671 946321 1. (CC) gcc options: -pthread -lpthread -O3 -march=native
Rodinia Test: OpenMP CFD Solver 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Seconds, Fewer Is Better Rodinia 2.4 Test: OpenMP CFD Solver no retpoline Retpoline Retpoline + GCC noretpoline 3 6 9 12 15 SE +/- 0.16, N = 3 SE +/- 0.15, N = 5 SE +/- 0.30, N = 6 SE +/- 0.04, N = 3 SE +/- 0.08, N = 3 SE +/- 0.11, N = 3 9.73 9.93 10.54 10.76 10.97 11.08 1. (CXX) g++ options: -O2 -lOpenCL
Compile Bench Test: Compile EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org MB/s, More Is Better Compile Bench 0.6 Test: Compile noretpoline Retpoline Retpoline + GCC no retpoline 400 800 1200 1600 2000 SE +/- 9.70, N = 3 SE +/- 4.84, N = 3 SE +/- 3.64, N = 3 SE +/- 62.71, N = 6 SE +/- 40.27, N = 6 SE +/- 7.13, N = 3 1696.87 1693.10 1691.37 1694.04 1637.02 1496.78
CacheBench Test: Read / Modify / Write 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org MB/s, More Is Better CacheBench Test: Read / Modify / Write no retpoline Retpoline Retpoline + GCC noretpoline 6K 12K 18K 24K 30K SE +/- 112.30, N = 3 SE +/- 271.49, N = 3 SE +/- 226.86, N = 3 SE +/- 5.63, N = 3 SE +/- 9.68, N = 3 SE +/- 12.91, N = 3 25777.79 25061.33 24928.44 22924.24 22918.96 22900.81 1. (CC) gcc options: -lrt
Rodinia Test: OpenMP LavaMD 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Seconds, Fewer Is Better Rodinia 2.4 Test: OpenMP LavaMD Retpoline + GCC no retpoline Retpoline noretpoline 7 14 21 28 35 SE +/- 0.08, N = 3 SE +/- 0.11, N = 3 SE +/- 0.16, N = 3 SE +/- 0.14, N = 3 SE +/- 0.08, N = 3 SE +/- 0.14, N = 3 28.43 28.50 28.70 31.65 31.86 31.87 1. (CXX) g++ options: -O2 -lOpenCL
C-Ray Total Time 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Seconds, Fewer Is Better C-Ray 1.1 Total Time Retpoline no retpoline Retpoline + GCC noretpoline 0.7875 1.575 2.3625 3.15 3.9375 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 3.15 3.16 3.18 3.45 3.48 3.50 1. (CC) gcc options: -lm -lpthread -O3
lzbench Test: Zstd 1 - Process: Compression 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org MB/s, More Is Better lzbench 2017-08-08 Test: Zstd 1 - Process: Compression Retpoline + GCC Retpoline no retpoline noretpoline 80 160 240 320 400 SE +/- 2.08, N = 3 SE +/- 2.52, N = 3 SE +/- 0.88, N = 3 SE +/- 1.33, N = 3 SE +/- 1.20, N = 3 SE +/- 1.33, N = 3 370 369 361 336 335 335 1. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3
libjpeg-turbo tjbench Test: Decompression Throughput 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Megapixels/sec, More Is Better libjpeg-turbo tjbench 1.5.1 Test: Decompression Throughput Retpoline Retpoline + GCC no retpoline noretpoline 30 60 90 120 150 SE +/- 1.27, N = 3 SE +/- 6.27, N = 6 SE +/- 5.29, N = 6 SE +/- 0.00, N = 3 SE +/- 0.07, N = 3 SE +/- 0.09, N = 3 154.02 146.51 145.81 140.99 140.87 140.83 1. (CC) gcc options: -O3 -lm
lzbench Test: Zstd 1 - Process: Decompression 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org MB/s, More Is Better lzbench 2017-08-08 Test: Zstd 1 - Process: Decompression Retpoline + GCC Retpoline no retpoline noretpoline 200 400 600 800 1000 SE +/- 3.00, N = 3 SE +/- 4.18, N = 3 SE +/- 4.58, N = 3 SE +/- 4.51, N = 3 SE +/- 4.33, N = 3 SE +/- 3.93, N = 3 983 982 975 911 911 910 1. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3
PostgreSQL pgbench Scaling: Buffer Test - Test: Normal Load - Mode: Read Only OpenBenchmarking.org TPS, More Is Better PostgreSQL pgbench 10.0 Scaling: Buffer Test - Test: Normal Load - Mode: Read Only no retpoline Retpoline Retpoline + GCC 130K 260K 390K 520K 650K SE +/- 2613.28, N = 3 SE +/- 1869.94, N = 3 SE +/- 4346.70, N = 3 599258.14 597483.03 577641.99 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -fPIC -lpgcommon -lpgport -lpthread -lrt -lcrypt -ldl -lm
Redis Test: GET 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Requests Per Second, More Is Better Redis 3.0.1 Test: GET Retpoline no retpoline Retpoline + GCC noretpoline 400K 800K 1200K 1600K 2000K SE +/- 96355.24, N = 6 SE +/- 50934.82, N = 6 SE +/- 80851.67, N = 6 SE +/- 2966.06, N = 3 SE +/- 27266.47, N = 3 SE +/- 28137.42, N = 6 1738340.06 1627717.71 1574731.42 1394712.79 1362255.92 1352327.02 1. (CC) gcc options: -ggdb -rdynamic -lm -pthread
Redis Test: LPUSH 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Requests Per Second, More Is Better Redis 3.0.1 Test: LPUSH Retpoline + GCC Retpoline no retpoline noretpoline 300K 600K 900K 1200K 1500K SE +/- 49927.23, N = 6 SE +/- 51060.27, N = 6 SE +/- 57810.96, N = 6 SE +/- 3427.20, N = 3 SE +/- 6595.02, N = 3 SE +/- 8338.77, N = 3 1454926.31 1382896.04 1357638.81 1147686.87 1123252.42 1117859.50 1. (CC) gcc options: -ggdb -rdynamic -lm -pthread
Redis Test: SADD 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Requests Per Second, More Is Better Redis 3.0.1 Test: SADD Retpoline + GCC no retpoline Retpoline noretpoline 300K 600K 900K 1200K 1500K SE +/- 62798.70, N = 6 SE +/- 68212.97, N = 6 SE +/- 95230.31, N = 6 SE +/- 51667.16, N = 6 SE +/- 7598.74, N = 3 SE +/- 19184.10, N = 4 1604656.60 1589703.69 1573307.83 1293172.50 1249051.08 1233207.56 1. (CC) gcc options: -ggdb -rdynamic -lm -pthread
Redis Test: LPOP EPYC 7601 2 x Xeon Gold 6138 OpenBenchmarking.org Requests Per Second, More Is Better Redis 3.0.1 Test: LPOP noretpoline Retpoline no retpoline Retpoline + GCC 300K 600K 900K 1200K 1500K SE +/- 1539.75, N = 3 SE +/- 61195.38, N = 6 SE +/- 11491.58, N = 3 SE +/- 77843.26, N = 6 SE +/- 36207.27, N = 6 SE +/- 25234.38, N = 6 1520530.25 1394465.79 1230732.38 1486377.46 1423431.25 1347143.44 1. (CC) gcc options: -ggdb -rdynamic -lm -pthread
Timed Linux Kernel Compilation Time To Compile 2 x Xeon Gold 6138 EPYC 7601 OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 4.13 Time To Compile no retpoline Retpoline Retpoline + GCC noretpoline 9 18 27 36 45 SE +/- 0.75, N = 6 SE +/- 0.77, N = 6 SE +/- 0.80, N = 6 SE +/- 0.68, N = 6 SE +/- 0.76, N = 6 SE +/- 0.63, N = 6 30.25 30.46 30.63 38.30 38.44 38.52
Phoronix Test Suite v10.8.4