Ubuntu 18.04 L1TF Foreshadow VM Testing Tests done from an Ubuntu 18.04 LTS x86_64 dual Xeon Gold 6138 Tyan server (the one frequently used in other Phoronix benchmarks). Host and VM using the Ubuntu packages including its patched 4.15 based kernel. Mitigation levels/tunables done after reboot each time from the bare metal system. VM was lone resource intensive process on the server and configured to access 80% of the host system's CPU resources, Tests by Michael Larabel for a future article.
HTML result view exported from: https://openbenchmarking.org/result/1808177-RA-XEONGOLD746&sro .
Ubuntu 18.04 L1TF Foreshadow VM Testing Processor Motherboard Chipset Memory Disk Graphics Audio Network OS Kernel Desktop Display Server Display Driver OpenGL Compiler File-System Screen Resolution System Layer Unmitigated Default Mitigation Always Flushing Full Mitigation 64 x Intel Xeon (Skylake IBRS) @ 2.00GHz (64 Cores) QEMU Standard PC (i440FX + PIIX 1996) (1.10.2-1ubuntu1 BIOS) Intel 440FX- 82441FX PMC 47104MB 162GB llvmpipe 46848MB QEMU Generic Red Hat Virtio device Ubuntu 18.04 4.15.0-29-generic (x86_64) GNOME Shell 3.28.2 X Server 1.19.6 modesetting 1.19.6 3.3 Mesa 18.0.5 (LLVM 6.0 256 bits) GCC 7.3.0 ext4 1024x768 KVM 4.15.0-32-generic (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-as=/usr/bin/x86_64-linux-gnu-as --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-ld=/usr/bin/x86_64-linux-gnu-ld --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v Disk Details - NONE / data=ordered,errors=remount-ro,relatime,rw Python Details - Python 2.7.15rc1 + Python 3.6.5 Security Details - Unmitigated: KPTI + __user pointer sanitization + Full generic retpoline IBPB IBRS_FW Protection - Default Mitigation: KPTI + __user pointer sanitization + Full generic retpoline IBPB IBRS_FW + PTE Inversion Protection - Always Flushing: KPTI + __user pointer sanitization + Full generic retpoline IBPB IBRS_FW + PTE Inversion Protection - Full Mitigation: KPTI + __user pointer sanitization + Full generic retpoline IBPB IBRS_FW + PTE Inversion Protection
Ubuntu 18.04 L1TF Foreshadow VM Testing sqlite: Timed SQLite Insertions fio: Rand Write - Linux AIO - No - Yes - 2MB - Default Test Directory fio: Seq Write - Linux AIO - No - Yes - 2MB - Default Test Directory iozone: 1MB - 4GB - Write Performance compilebench: Compile compilebench: Initial Create compilebench: Read Compiled Tree sockperf: Throughput sockperf: Latency Ping Pong sockperf: Latency Under Load ttsiod-renderer: Phong Rendering With Soft-Shadow Mapping ebizzy: build-linux-kernel: Time To Compile build-llvm: Time To Compile compress-zstd: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19 hackbench: 4 - Thread hackbench: 8 - Thread hackbench: 16 - Thread hackbench: 4 - Process hackbench: 8 - Process hackbench: 16 - Process openssl: RSA 4096-bit Performance mysqlslap: 8 pgbench: Buffer Test - Normal Load - Read Only pgbench: Buffer Test - Normal Load - Read Write redis: LPOP redis: SADD redis: LPUSH redis: GET redis: SET stress-ng: Semaphores stress-ng: Matrix Math stress-ng: Vector Math stress-ng: Socket Activity stress-ng: Context Switching stress-ng: System V Message Passing sysbench: Memory sysbench: CPU nginx: Static Web Page Serving apache: Static Web Page Serving apache-siege: 50 git: Time To Complete Common Git Commands fio: Rand Read - Linux AIO - No - Yes - 2MB - Default Test Directory fio: Seq Read - Linux AIO - No - Yes - 2MB - Default Test Directory fio: Rand Read - POSIX AIO - No - Yes - 2MB - Default Test Directory fio: Rand Write - POSIX AIO - No - Yes - 2MB - Default Test Directory fio: Seq Read - POSIX AIO - No - Yes - 2MB - Default Test Directory fio: Seq Write - POSIX AIO - No - Yes - 2MB - Default Test Directory gimp: unsharp-mask gimp: resize gimp: rotate gimp: auto-levels Unmitigated Default Mitigation Always Flushing Full Mitigation 61.99 1972 1318 309.22 1402.16 462.67 2234.53 464697 7.14 45.94 660.20 48618 33.85 195.49 144.83 9.42 18.87 36.97 9.49 20.11 37.04 7915.97 245 394876.94 5152.08 2068611.85 1878780.17 1297688.50 2026228.33 1463285.46 557628.55 168491.90 60351.43 13880.53 12590168.74 12135778.98 10438083.96 51572.28 27495.41 24525.21 14060.58 7.35 12100 11367 1095 766 1242 712 54.01 11.61 24.11 25.12 63.28 2153 1444 311.44 1450.63 444.94 2233.48 489081 7.80 46.84 651.56 44835 33.73 203.77 140.34 9.51 19.30 38.03 9.59 19.81 36.59 7879.13 244 387186.23 5162.44 1840381.56 1746629.66 1431348.01 1824211.40 1531909.19 599437.06 169670.38 60406.78 13163.12 12482470.05 12276978.98 10535022.39 51480.55 26983.21 24714.34 13700.33 7.28 11311 11400 1115 830 1247 737 54.01 11.25 24.20 24.61 62.71 2132 1432 314.77 1408.53 451.08 2237.12 468362 8.70 49.14 652.36 53068 33.64 199.40 145.58 10.58 21.76 43.36 10.78 24.19 43.53 7833.37 237 376404.54 5126.91 1443664.99 1669834.00 1418073.81 1762582.37 1421644.67 581171.78 171874.04 60504.84 12992.89 12370237.98 12192475.24 10510051.38 51558.10 26388.08 22456.19 13502.71 7.27 10767 10333 1191 788 1203 726 56.35 11.83 24.77 25.32 62.49 2044 1219 304.39 1360.12 411.13 2178.71 477577 6.90 53.28 206.30 76020 38.00 219.11 148.09 21.72 30.39 54.12 21.91 31.94 61.49 7607.07 235 263895.66 4859.91 1397732.79 1625370.68 1405706.45 1789732.77 1501864.62 700793.02 136088.10 45524.09 12833.20 11050870.69 7914016.08 7267419.40 36587.25 26712.57 22023.62 8015.48 7.30 10533 11033 1180 761 1259 740 55.82 12.17 25.28 26.17 OpenBenchmarking.org
SQLite Timed SQLite Insertions OpenBenchmarking.org Seconds, Fewer Is Better SQLite 3.22 Timed SQLite Insertions Always Flushing Default Mitigation Full Mitigation Unmitigated 14 28 42 56 70 SE +/- 1.01, N = 3 SE +/- 0.78, N = 3 SE +/- 0.46, N = 3 SE +/- 0.04, N = 3 62.71 63.28 62.49 61.99 1. (CC) gcc options: -O2 -lz -ldl -lpthread
Flexible IO Tester Type: Random Write - IO Engine: Linux AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 3.1 Type: Random Write - IO Engine: Linux AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Always Flushing Default Mitigation Full Mitigation Unmitigated 500 1000 1500 2000 2500 SE +/- 31.73, N = 12 SE +/- 34.60, N = 9 SE +/- 4.36, N = 3 SE +/- 38.30, N = 6 2132 2153 2044 1972 1. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -ll -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Write - IO Engine: Linux AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 3.1 Type: Sequential Write - IO Engine: Linux AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Always Flushing Default Mitigation Full Mitigation Unmitigated 300 600 900 1200 1500 SE +/- 57.61, N = 9 SE +/- 75.95, N = 9 SE +/- 11.29, N = 3 SE +/- 53.63, N = 6 1432 1444 1219 1318 1. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -ll -lrt -laio -lz -lm -lpthread -ldl
IOzone Record Size: 1MB - File Size: 4GB - Disk Test: Write Performance OpenBenchmarking.org MB/s, More Is Better IOzone 3.465 Record Size: 1MB - File Size: 4GB - Disk Test: Write Performance Always Flushing Default Mitigation Full Mitigation Unmitigated 70 140 210 280 350 SE +/- 1.21, N = 3 SE +/- 1.19, N = 3 SE +/- 4.45, N = 3 SE +/- 4.67, N = 3 314.77 311.44 304.39 309.22 1. (CC) gcc options: -O3
Compile Bench Test: Compile OpenBenchmarking.org MB/s, More Is Better Compile Bench 0.6 Test: Compile Always Flushing Default Mitigation Full Mitigation Unmitigated 300 600 900 1200 1500 SE +/- 24.42, N = 3 SE +/- 23.36, N = 4 SE +/- 26.60, N = 3 SE +/- 36.69, N = 6 1408.53 1450.63 1360.12 1402.16
Compile Bench Test: Initial Create OpenBenchmarking.org MB/s, More Is Better Compile Bench 0.6 Test: Initial Create Always Flushing Default Mitigation Full Mitigation Unmitigated 100 200 300 400 500 SE +/- 10.50, N = 3 SE +/- 1.76, N = 3 SE +/- 17.84, N = 3 SE +/- 5.89, N = 3 451.08 444.94 411.13 462.67
Compile Bench Test: Read Compiled Tree OpenBenchmarking.org MB/s, More Is Better Compile Bench 0.6 Test: Read Compiled Tree Always Flushing Default Mitigation Full Mitigation Unmitigated 500 1000 1500 2000 2500 SE +/- 21.33, N = 3 SE +/- 40.11, N = 3 SE +/- 33.25, N = 3 SE +/- 8.33, N = 3 2237.12 2233.48 2178.71 2234.53
Sockperf Test: Throughput OpenBenchmarking.org Messages Per Second, More Is Better Sockperf 3.4 Test: Throughput Always Flushing Default Mitigation Full Mitigation Unmitigated 100K 200K 300K 400K 500K SE +/- 15140.22, N = 15 SE +/- 10160.93, N = 20 SE +/- 15206.71, N = 20 SE +/- 22194.18, N = 10 468362 489081 477577 464697 1. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread
Sockperf Test: Latency Ping Pong OpenBenchmarking.org usec, Fewer Is Better Sockperf 3.4 Test: Latency Ping Pong Always Flushing Default Mitigation Full Mitigation Unmitigated 2 4 6 8 10 SE +/- 0.37, N = 20 SE +/- 0.34, N = 15 SE +/- 0.63, N = 15 SE +/- 0.59, N = 10 8.70 7.80 6.90 7.14 1. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread
Sockperf Test: Latency Under Load OpenBenchmarking.org usec, Fewer Is Better Sockperf 3.4 Test: Latency Under Load Always Flushing Default Mitigation Full Mitigation Unmitigated 12 24 36 48 60 SE +/- 4.00, N = 20 SE +/- 4.68, N = 15 SE +/- 3.21, N = 16 SE +/- 3.40, N = 10 49.14 46.84 53.28 45.94 1. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread
TTSIOD 3D Renderer Phong Rendering With Soft-Shadow Mapping OpenBenchmarking.org FPS, More Is Better TTSIOD 3D Renderer 2.3b Phong Rendering With Soft-Shadow Mapping Always Flushing Default Mitigation Full Mitigation Unmitigated 140 280 420 560 700 SE +/- 10.57, N = 3 SE +/- 11.07, N = 4 SE +/- 0.31, N = 3 SE +/- 6.92, N = 3 652.36 651.56 206.30 660.20 1. (CXX) g++ options: -O3 -fomit-frame-pointer -ffast-math -mtune=native -flto -msse -mrecip -mfpmath=sse -msse2 -mssse3 -lSDL -fopenmp -fwhole-program -lstdc++
ebizzy OpenBenchmarking.org Records/s, More Is Better ebizzy 0.3 Always Flushing Default Mitigation Full Mitigation Unmitigated 16K 32K 48K 64K 80K SE +/- 1018.90, N = 3 SE +/- 404.70, N = 3 SE +/- 1286.76, N = 4 SE +/- 429.33, N = 3 53068 44835 76020 48618 1. (CC) gcc options: -pthread -lpthread -O3 -march=native
Timed Linux Kernel Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 4.13 Time To Compile Always Flushing Default Mitigation Full Mitigation Unmitigated 9 18 27 36 45 SE +/- 0.37, N = 12 SE +/- 0.53, N = 9 SE +/- 0.45, N = 12 SE +/- 0.81, N = 6 33.64 33.73 38.00 33.85
Timed LLVM Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed LLVM Compilation 6.0.1 Time To Compile Always Flushing Default Mitigation Full Mitigation Unmitigated 50 100 150 200 250 199.40 203.77 219.11 195.49
Zstd Compression Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19 OpenBenchmarking.org Seconds, Fewer Is Better Zstd Compression 1.3.4 Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19 Always Flushing Default Mitigation Full Mitigation Unmitigated 30 60 90 120 150 SE +/- 1.23, N = 3 SE +/- 1.05, N = 3 SE +/- 0.47, N = 3 SE +/- 1.55, N = 3 145.58 140.34 148.09 144.83 1. (CC) gcc options: -O3 -pthread -lz -llzma
Hackbench Count: 4 - Type: Thread OpenBenchmarking.org Seconds, Fewer Is Better Hackbench Count: 4 - Type: Thread Always Flushing Default Mitigation Full Mitigation Unmitigated 5 10 15 20 25 SE +/- 0.07, N = 3 SE +/- 0.03, N = 3 SE +/- 0.32, N = 3 SE +/- 0.07, N = 3 10.58 9.51 21.72 9.42 1. (CC) gcc options: -lpthread
Hackbench Count: 8 - Type: Thread OpenBenchmarking.org Seconds, Fewer Is Better Hackbench Count: 8 - Type: Thread Always Flushing Default Mitigation Full Mitigation Unmitigated 7 14 21 28 35 SE +/- 0.19, N = 3 SE +/- 0.18, N = 3 SE +/- 0.18, N = 3 SE +/- 0.09, N = 3 21.76 19.30 30.39 18.87 1. (CC) gcc options: -lpthread
Hackbench Count: 16 - Type: Thread OpenBenchmarking.org Seconds, Fewer Is Better Hackbench Count: 16 - Type: Thread Always Flushing Default Mitigation Full Mitigation Unmitigated 12 24 36 48 60 SE +/- 0.71, N = 3 SE +/- 0.25, N = 3 SE +/- 0.78, N = 4 SE +/- 0.26, N = 3 43.36 38.03 54.12 36.97 1. (CC) gcc options: -lpthread
Hackbench Count: 4 - Type: Process OpenBenchmarking.org Seconds, Fewer Is Better Hackbench Count: 4 - Type: Process Always Flushing Default Mitigation Full Mitigation Unmitigated 5 10 15 20 25 SE +/- 0.05, N = 3 SE +/- 0.08, N = 3 SE +/- 0.11, N = 3 SE +/- 0.09, N = 3 10.78 9.59 21.91 9.49 1. (CC) gcc options: -lpthread
Hackbench Count: 8 - Type: Process OpenBenchmarking.org Seconds, Fewer Is Better Hackbench Count: 8 - Type: Process Always Flushing Default Mitigation Full Mitigation Unmitigated 7 14 21 28 35 SE +/- 0.37, N = 4 SE +/- 0.20, N = 3 SE +/- 0.22, N = 3 SE +/- 0.07, N = 3 24.19 19.81 31.94 20.11 1. (CC) gcc options: -lpthread
Hackbench Count: 16 - Type: Process OpenBenchmarking.org Seconds, Fewer Is Better Hackbench Count: 16 - Type: Process Always Flushing Default Mitigation Full Mitigation Unmitigated 14 28 42 56 70 SE +/- 0.51, N = 3 SE +/- 0.12, N = 3 SE +/- 1.05, N = 3 SE +/- 0.27, N = 3 43.53 36.59 61.49 37.04 1. (CC) gcc options: -lpthread
OpenSSL RSA 4096-bit Performance OpenBenchmarking.org Signs Per Second, More Is Better OpenSSL 1.1.0f RSA 4096-bit Performance Always Flushing Default Mitigation Full Mitigation Unmitigated 2K 4K 6K 8K 10K SE +/- 31.06, N = 3 SE +/- 32.73, N = 3 SE +/- 79.42, N = 3 SE +/- 14.57, N = 3 7833.37 7879.13 7607.07 7915.97 1. (CC) gcc options: -O3 -pthread -m64 -lssl -lcrypto -ldl
MariaDB Clients: 8 OpenBenchmarking.org Queries Per Second, More Is Better MariaDB 10.3.8 Clients: 8 Always Flushing Default Mitigation Full Mitigation Unmitigated 50 100 150 200 250 SE +/- 1.95, N = 3 SE +/- 5.43, N = 3 SE +/- 0.82, N = 3 SE +/- 1.89, N = 3 237 244 235 245 1. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -llzma -laio -lz -lm -lpcre -lcrypt -lssl -lcrypto -ldl
PostgreSQL pgbench Scaling: Buffer Test - Test: Normal Load - Mode: Read Only OpenBenchmarking.org TPS, More Is Better PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Normal Load - Mode: Read Only Always Flushing Default Mitigation Full Mitigation Unmitigated 80K 160K 240K 320K 400K SE +/- 2604.84, N = 3 SE +/- 2783.33, N = 3 SE +/- 1013.98, N = 3 SE +/- 2565.21, N = 3 376404.54 387186.23 263895.66 394876.94 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm
PostgreSQL pgbench Scaling: Buffer Test - Test: Normal Load - Mode: Read Write OpenBenchmarking.org TPS, More Is Better PostgreSQL pgbench 10.3 Scaling: Buffer Test - Test: Normal Load - Mode: Read Write Always Flushing Default Mitigation Full Mitigation Unmitigated 1100 2200 3300 4400 5500 SE +/- 3.70, N = 3 SE +/- 32.52, N = 3 SE +/- 70.66, N = 3 SE +/- 15.96, N = 3 5126.91 5162.44 4859.91 5152.08 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm
Redis Test: LPOP OpenBenchmarking.org Requests Per Second, More Is Better Redis 4.0.8 Test: LPOP Always Flushing Default Mitigation Full Mitigation Unmitigated 400K 800K 1200K 1600K 2000K SE +/- 53543.04, N = 10 SE +/- 101684.52, N = 12 SE +/- 35832.53, N = 12 SE +/- 137342.99, N = 6 1443664.99 1840381.56 1397732.79 2068611.85 1. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread
Redis Test: SADD OpenBenchmarking.org Requests Per Second, More Is Better Redis 4.0.8 Test: SADD Always Flushing Default Mitigation Full Mitigation Unmitigated 400K 800K 1200K 1600K 2000K SE +/- 61306.24, N = 12 SE +/- 69330.61, N = 9 SE +/- 49376.90, N = 12 SE +/- 85165.61, N = 6 1669834.00 1746629.66 1625370.68 1878780.17 1. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread
Redis Test: LPUSH OpenBenchmarking.org Requests Per Second, More Is Better Redis 4.0.8 Test: LPUSH Always Flushing Default Mitigation Full Mitigation Unmitigated 300K 600K 900K 1200K 1500K SE +/- 53016.95, N = 12 SE +/- 39118.19, N = 9 SE +/- 41599.43, N = 12 SE +/- 20709.64, N = 3 1418073.81 1431348.01 1405706.45 1297688.50 1. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread
Redis Test: GET OpenBenchmarking.org Requests Per Second, More Is Better Redis 4.0.8 Test: GET Always Flushing Default Mitigation Full Mitigation Unmitigated 400K 800K 1200K 1600K 2000K SE +/- 72106.90, N = 12 SE +/- 92589.67, N = 9 SE +/- 48565.17, N = 12 SE +/- 87966.15, N = 6 1762582.37 1824211.40 1789732.77 2026228.33 1. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread
Redis Test: SET OpenBenchmarking.org Requests Per Second, More Is Better Redis 4.0.8 Test: SET Always Flushing Default Mitigation Full Mitigation Unmitigated 300K 600K 900K 1200K 1500K SE +/- 55957.00, N = 12 SE +/- 60083.85, N = 9 SE +/- 37254.96, N = 12 SE +/- 88955.87, N = 6 1421644.67 1531909.19 1501864.62 1463285.46 1. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread
Stress-NG Test: Semaphores OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.07.26 Test: Semaphores Always Flushing Default Mitigation Full Mitigation Unmitigated 150K 300K 450K 600K 750K SE +/- 584.43, N = 3 SE +/- 6344.60, N = 3 SE +/- 10838.51, N = 5 SE +/- 11434.26, N = 6 581171.78 599437.06 700793.02 557628.55 1. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
Stress-NG Test: Matrix Math OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.07.26 Test: Matrix Math Always Flushing Default Mitigation Full Mitigation Unmitigated 40K 80K 120K 160K 200K SE +/- 1078.42, N = 3 SE +/- 1660.13, N = 3 SE +/- 982.93, N = 3 SE +/- 2677.89, N = 3 171874.04 169670.38 136088.10 168491.90 1. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
Stress-NG Test: Vector Math OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.07.26 Test: Vector Math Always Flushing Default Mitigation Full Mitigation Unmitigated 13K 26K 39K 52K 65K SE +/- 30.21, N = 3 SE +/- 298.54, N = 3 SE +/- 137.57, N = 3 SE +/- 186.86, N = 3 60504.84 60406.78 45524.09 60351.43 1. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
Stress-NG Test: Socket Activity OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.07.26 Test: Socket Activity Always Flushing Default Mitigation Full Mitigation Unmitigated 3K 6K 9K 12K 15K SE +/- 98.54, N = 3 SE +/- 47.46, N = 3 SE +/- 40.44, N = 3 SE +/- 90.47, N = 3 12992.89 13163.12 12833.20 13880.53 1. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
Stress-NG Test: Context Switching OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.07.26 Test: Context Switching Always Flushing Default Mitigation Full Mitigation Unmitigated 3M 6M 9M 12M 15M SE +/- 141287.57, N = 3 SE +/- 12082.59, N = 3 SE +/- 155557.89, N = 3 SE +/- 177083.84, N = 3 12370237.98 12482470.05 11050870.69 12590168.74 1. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
Stress-NG Test: System V Message Passing OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.07.26 Test: System V Message Passing Always Flushing Default Mitigation Full Mitigation Unmitigated 3M 6M 9M 12M 15M SE +/- 113733.08, N = 3 SE +/- 95795.42, N = 3 SE +/- 86447.80, N = 3 SE +/- 188941.31, N = 3 12192475.24 12276978.98 7914016.08 12135778.98 1. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
Sysbench Test: Memory OpenBenchmarking.org Events Per Second, More Is Better Sysbench 2018-07-28 Test: Memory Always Flushing Default Mitigation Full Mitigation Unmitigated 2M 4M 6M 8M 10M SE +/- 139686.71, N = 3 SE +/- 151324.74, N = 3 SE +/- 114162.45, N = 3 SE +/- 73702.36, N = 3 10510051.38 10535022.39 7267419.40 10438083.96 1. (CC) gcc options: -pthread -O3 -funroll-loops -ggdb3 -march=core2 -rdynamic -ldl -laio -lm
Sysbench Test: CPU OpenBenchmarking.org Events Per Second, More Is Better Sysbench 2018-07-28 Test: CPU Always Flushing Default Mitigation Full Mitigation Unmitigated 11K 22K 33K 44K 55K SE +/- 36.69, N = 3 SE +/- 24.09, N = 3 SE +/- 20.94, N = 3 SE +/- 52.27, N = 3 51558.10 51480.55 36587.25 51572.28 1. (CC) gcc options: -pthread -O3 -funroll-loops -ggdb3 -march=core2 -rdynamic -ldl -laio -lm
NGINX Benchmark Static Web Page Serving OpenBenchmarking.org Requests Per Second, More Is Better NGINX Benchmark 1.9.9 Static Web Page Serving Always Flushing Default Mitigation Full Mitigation Unmitigated 6K 12K 18K 24K 30K SE +/- 411.39, N = 5 SE +/- 200.40, N = 3 SE +/- 90.18, N = 3 SE +/- 248.68, N = 3 26388.08 26983.21 26712.57 27495.41 1. (CC) gcc options: -lpthread -lcrypt -lcrypto -lz -O3 -march=native
Apache Benchmark Static Web Page Serving OpenBenchmarking.org Requests Per Second, More Is Better Apache Benchmark 2.4.29 Static Web Page Serving Always Flushing Default Mitigation Full Mitigation Unmitigated 5K 10K 15K 20K 25K SE +/- 292.86, N = 7 SE +/- 361.69, N = 9 SE +/- 349.24, N = 12 SE +/- 471.64, N = 3 22456.19 24714.34 22023.62 24525.21 1. (CC) gcc options: -shared -fPIC -O2 -pthread
Apache Siege Concurrent Users: 50 OpenBenchmarking.org Transactions Per Second, More Is Better Apache Siege 2.4.29 Concurrent Users: 50 Always Flushing Default Mitigation Full Mitigation Unmitigated 3K 6K 9K 12K 15K 13502.71 13700.33 8015.48 14060.58 1. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto
Git Time To Complete Common Git Commands OpenBenchmarking.org Seconds, Fewer Is Better Git Time To Complete Common Git Commands Always Flushing Default Mitigation Full Mitigation Unmitigated 2 4 6 8 10 SE +/- 0.06, N = 3 SE +/- 0.10, N = 3 SE +/- 0.11, N = 3 SE +/- 0.10, N = 3 7.27 7.28 7.30 7.35 1. git version 2.17.1
Flexible IO Tester Type: Random Read - IO Engine: Linux AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 3.1 Type: Random Read - IO Engine: Linux AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Always Flushing Default Mitigation Full Mitigation Unmitigated 3K 6K 9K 12K 15K SE +/- 33.33, N = 3 SE +/- 246.33, N = 9 SE +/- 133.33, N = 3 SE +/- 272.03, N = 6 10767 11311 10533 12100 1. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -ll -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Read - IO Engine: Linux AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory OpenBenchmarking.org IOPS, More Is Better Flexible IO Tester 3.1 Type: Sequential Read - IO Engine: Linux AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Always Flushing Default Mitigation Full Mitigation Unmitigated 2K 4K 6K 8K 10K SE +/- 145.30, N = 3 SE +/- 303.68, N = 9 SE +/- 133.33, N = 3 SE +/- 176.38, N = 3 10333 11400 11033 11367 1. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -ll -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Random Read - IO Engine: POSIX AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 3.1 Type: Random Read - IO Engine: POSIX AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Always Flushing Default Mitigation Full Mitigation Unmitigated 300 600 900 1200 1500 SE +/- 5.55, N = 3 SE +/- 17.38, N = 3 SE +/- 27.55, N = 12 SE +/- 8.37, N = 3 1191 1115 1180 1095 1. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -ll -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Random Write - IO Engine: POSIX AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 3.1 Type: Random Write - IO Engine: POSIX AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Always Flushing Default Mitigation Full Mitigation Unmitigated 200 400 600 800 1000 SE +/- 10.82, N = 3 SE +/- 8.08, N = 3 SE +/- 9.93, N = 12 SE +/- 11.97, N = 6 788 830 761 766 1. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -ll -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Read - IO Engine: POSIX AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 3.1 Type: Sequential Read - IO Engine: POSIX AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Always Flushing Default Mitigation Full Mitigation Unmitigated 300 600 900 1200 1500 SE +/- 25.02, N = 12 SE +/- 35.68, N = 9 SE +/- 19.30, N = 12 SE +/- 11.57, N = 3 1203 1247 1259 1242 1. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -ll -lrt -laio -lz -lm -lpthread -ldl
Flexible IO Tester Type: Sequential Write - IO Engine: POSIX AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory OpenBenchmarking.org MB/s, More Is Better Flexible IO Tester 3.1 Type: Sequential Write - IO Engine: POSIX AIO - Buffered: No - Direct: Yes - Block Size: 2MB - Disk Target: Default Test Directory Always Flushing Default Mitigation Full Mitigation Unmitigated 160 320 480 640 800 SE +/- 12.34, N = 3 SE +/- 20.23, N = 9 SE +/- 8.57, N = 12 SE +/- 11.85, N = 3 726 737 740 712 1. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -ll -lrt -laio -lz -lm -lpthread -ldl
GIMP Test: unsharp-mask OpenBenchmarking.org Seconds, Fewer Is Better GIMP 2.8.22 Test: unsharp-mask Always Flushing Default Mitigation Full Mitigation Unmitigated 13 26 39 52 65 SE +/- 0.28, N = 3 SE +/- 0.34, N = 3 SE +/- 0.49, N = 3 SE +/- 0.35, N = 3 56.35 54.01 55.82 54.01
GIMP Test: resize OpenBenchmarking.org Seconds, Fewer Is Better GIMP 2.8.22 Test: resize Always Flushing Default Mitigation Full Mitigation Unmitigated 3 6 9 12 15 SE +/- 0.09, N = 3 SE +/- 0.07, N = 3 SE +/- 0.05, N = 3 SE +/- 0.23, N = 3 11.83 11.25 12.17 11.61
GIMP Test: rotate OpenBenchmarking.org Seconds, Fewer Is Better GIMP 2.8.22 Test: rotate Always Flushing Default Mitigation Full Mitigation Unmitigated 6 12 18 24 30 SE +/- 0.19, N = 3 SE +/- 0.13, N = 3 SE +/- 0.34, N = 3 SE +/- 0.16, N = 3 24.77 24.20 25.28 24.11
GIMP Test: auto-levels OpenBenchmarking.org Seconds, Fewer Is Better GIMP 2.8.22 Test: auto-levels Always Flushing Default Mitigation Full Mitigation Unmitigated 6 12 18 24 30 SE +/- 0.06, N = 3 SE +/- 0.29, N = 3 SE +/- 0.07, N = 3 SE +/- 0.24, N = 3 25.32 24.61 26.17 25.12
Phoronix Test Suite v10.8.4