ncnn epyc

2 x AMD EPYC 7F72 24-Core testing with a Supermicro H11DSi-NT v2.00 (2.1 BIOS) and ASPEED on Ubuntu 20.10 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2012193-HA-NCNNEPYC146
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
1
December 19 2020
  10 Hours, 45 Minutes
2
December 19 2020
  10 Hours, 15 Minutes
3
December 19 2020
  10 Hours, 27 Minutes
Invert Hiding All Results Option
  10 Hours, 29 Minutes

Only show results where is faster than
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):


ncnn epycProcessorMotherboardChipsetMemoryDiskGraphicsMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLCompilerFile-SystemScreen Resolution1232 x AMD EPYC 7F72 24-Core @ 3.20GHz (48 Cores / 96 Threads)Supermicro H11DSi-NT v2.00 (2.1 BIOS)AMD Starship/Matisse126GB1000GB Western Digital WD_BLACK SN850 1TBASPEEDVE2282 x Intel 10G X550TUbuntu 20.105.8.0-29-generic (x86_64)GNOME Shell 3.38.1X Server 1.20.9modesetting 1.20.94.5 Mesa 20.2.1 (LLVM 11.0.0 256 bits)GCC 10.2.0ext41920x1080OpenBenchmarking.orgCompiler 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++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-JvwpWM/gcc-10-10.2.0/debian/tmp-gcn/usr,hsa --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=auto --with-tune=generic --without-cuda-driver -v Processor Details- Scaling Governor: acpi-cpufreq ondemand (Boost: Enabled) - CPU Microcode: 0x8301034 Security Details- itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected

123Result OverviewPhoronix Test Suite100%109%118%127%136%NCNNNCNNNCNNNCNNNCNNNCNNNCNNNCNNNCNNNCNNNCNNNCNNNCNNNCNNNCNNCPU-v3-v3 - mobilenet-v3CPU - resnet18CPU - mnasnetCPU - shufflenet-v2CPU - mobilenetCPU - alexnetCPU - blazefaceCPU - efficientnet-b0CPU - resnet50CPU - yolov4-tinyCPU - vgg16CPU - googlenetCPU - regnety_400mCPU - squeezenet_ssdCPU-v2-v2 - mobilenet-v2

ncnn epycncnn: CPU - mobilenetncnn: CPU-v2-v2 - mobilenet-v2ncnn: CPU-v3-v3 - mobilenet-v3ncnn: CPU - shufflenet-v2ncnn: CPU - mnasnetncnn: CPU - efficientnet-b0ncnn: CPU - blazefacencnn: CPU - googlenetncnn: CPU - vgg16ncnn: CPU - resnet18ncnn: CPU - alexnetncnn: CPU - resnet50ncnn: CPU - yolov4-tinyncnn: CPU - squeezenet_ssdncnn: CPU - regnety_400m12360.3238.2247.7835.7435.1845.1015.6259.0673.3528.5411.3952.0849.0345.92193.2955.2238.1635.0732.6036.6348.9817.0255.9469.4324.0411.1950.1646.1345.85186.5861.3137.2336.3031.6041.7144.9615.8656.2267.9027.1512.2254.5545.2847.34184.95OpenBenchmarking.org

NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: mobilenet1231428425670SE +/- 2.86, N = 12SE +/- 3.18, N = 12SE +/- 4.95, N = 1260.3255.2261.31MIN: 30.39 / MAX: 248.99MIN: 34.09 / MAX: 278.82MIN: 34.14 / MAX: 1559.21. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: mobilenet1231224364860Min: 42.35 / Avg: 60.32 / Max: 74.82Min: 41.01 / Avg: 55.22 / Max: 73.28Min: 42.84 / Avg: 61.31 / Max: 92.991. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU-v2-v2 - Model: mobilenet-v2123918273645SE +/- 3.05, N = 12SE +/- 3.17, N = 12SE +/- 2.74, N = 1238.2238.1637.23MIN: 17.72 / MAX: 2514.73MIN: 17.73 / MAX: 2062.34MIN: 17.39 / MAX: 2908.091. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU-v2-v2 - Model: mobilenet-v2123816243240Min: 24.95 / Avg: 38.22 / Max: 55.97Min: 22.63 / Avg: 38.16 / Max: 63.14Min: 26.58 / Avg: 37.23 / Max: 54.51. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU-v3-v3 - Model: mobilenet-v31231122334455SE +/- 7.60, N = 12SE +/- 2.95, N = 12SE +/- 2.21, N = 1247.7835.0736.30MIN: 17.09 / MAX: 3569.21MIN: 16.82 / MAX: 2974.32MIN: 17.79 / MAX: 1511.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU-v3-v3 - Model: mobilenet-v31231020304050Min: 23.1 / Avg: 47.78 / Max: 115.99Min: 23.29 / Avg: 35.07 / Max: 61.36Min: 23.84 / Avg: 36.3 / Max: 51.411. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: shufflenet-v2123816243240SE +/- 3.66, N = 12SE +/- 2.54, N = 12SE +/- 2.73, N = 1235.7432.6031.60MIN: 17.04 / MAX: 2368.5MIN: 16.8 / MAX: 2388.01MIN: 18 / MAX: 2374.381. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: shufflenet-v2123816243240Min: 21.1 / Avg: 35.74 / Max: 70.18Min: 20.95 / Avg: 32.6 / Max: 54.16Min: 24.48 / Avg: 31.6 / Max: 59.291. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: mnasnet1231020304050SE +/- 3.22, N = 12SE +/- 2.33, N = 12SE +/- 8.15, N = 1235.1836.6341.71MIN: 16.4 / MAX: 554.91MIN: 16.89 / MAX: 504.39MIN: 17.49 / MAX: 2903.421. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: mnasnet123918273645Min: 19.53 / Avg: 35.18 / Max: 55.91Min: 27.5 / Avg: 36.63 / Max: 52.19Min: 21.3 / Avg: 41.71 / Max: 127.621. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: efficientnet-b01231122334455SE +/- 2.06, N = 12SE +/- 3.99, N = 12SE +/- 2.05, N = 1245.1048.9844.96MIN: 25.26 / MAX: 795.04MIN: 25.21 / MAX: 2676.06MIN: 24.8 / MAX: 1866.521. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: efficientnet-b01231020304050Min: 31.32 / Avg: 45.1 / Max: 52.46Min: 33.86 / Avg: 48.98 / Max: 78.34Min: 32.84 / Avg: 44.96 / Max: 54.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: blazeface12348121620SE +/- 0.43, N = 12SE +/- 1.29, N = 12SE +/- 0.72, N = 1215.6217.0215.86MIN: 9.96 / MAX: 86.56MIN: 9.89 / MAX: 1590.16MIN: 8.96 / MAX: 103.721. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: blazeface12348121620Min: 12.91 / Avg: 15.62 / Max: 18.3Min: 12.45 / Avg: 17.02 / Max: 29.44Min: 11.97 / Avg: 15.86 / Max: 21.461. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: googlenet1231326395265SE +/- 4.83, N = 12SE +/- 3.97, N = 12SE +/- 3.43, N = 1259.0655.9456.22MIN: 33.35 / MAX: 3976.49MIN: 30.92 / MAX: 3251.16MIN: 32.53 / MAX: 913.021. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: googlenet1231224364860Min: 36.81 / Avg: 59.06 / Max: 100.64Min: 33.35 / Avg: 55.94 / Max: 78.71Min: 39.18 / Avg: 56.22 / Max: 76.691. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: vgg161231632486480SE +/- 3.59, N = 12SE +/- 3.89, N = 12SE +/- 2.05, N = 1273.3569.4367.90MIN: 48.46 / MAX: 746.27MIN: 46.3 / MAX: 1136.73MIN: 46.51 / MAX: 790.071. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: vgg161231428425670Min: 55.08 / Avg: 73.35 / Max: 95.96Min: 53.8 / Avg: 69.43 / Max: 102.54Min: 55.3 / Avg: 67.9 / Max: 77.231. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: resnet18123714212835SE +/- 1.97, N = 12SE +/- 1.25, N = 12SE +/- 1.71, N = 1228.5424.0427.15MIN: 17.8 / MAX: 389.05MIN: 16.52 / MAX: 212.63MIN: 17.39 / MAX: 140.291. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: resnet18123612182430Min: 22.33 / Avg: 28.54 / Max: 46.31Min: 19.28 / Avg: 24.04 / Max: 34.98Min: 19.7 / Avg: 27.15 / Max: 40.281. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: alexnet1233691215SE +/- 0.65, N = 12SE +/- 0.57, N = 12SE +/- 0.57, N = 1211.3911.1912.22MIN: 8.6 / MAX: 110.34MIN: 8.26 / MAX: 356.88MIN: 8.32 / MAX: 179.051. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: alexnet12348121620Min: 9.53 / Avg: 11.39 / Max: 17.78Min: 9.36 / Avg: 11.19 / Max: 15.58Min: 10.03 / Avg: 12.22 / Max: 16.311. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: resnet501231224364860SE +/- 1.69, N = 12SE +/- 1.21, N = 12SE +/- 2.31, N = 1252.0850.1654.55MIN: 37.4 / MAX: 743.74MIN: 39.64 / MAX: 833.08MIN: 38.06 / MAX: 1626.461. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: resnet501231122334455Min: 44.66 / Avg: 52.08 / Max: 65.42Min: 43.78 / Avg: 50.16 / Max: 58.86Min: 42.85 / Avg: 54.55 / Max: 68.821. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: yolov4-tiny1231122334455SE +/- 1.97, N = 12SE +/- 1.06, N = 12SE +/- 1.01, N = 1249.0346.1345.28MIN: 36.2 / MAX: 1512.73MIN: 35.97 / MAX: 389.12MIN: 35.53 / MAX: 127.571. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: yolov4-tiny1231020304050Min: 44.19 / Avg: 49.03 / Max: 67.26Min: 40.06 / Avg: 46.13 / Max: 52.22Min: 39.95 / Avg: 45.28 / Max: 53.411. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: squeezenet_ssd1231122334455SE +/- 1.37, N = 12SE +/- 0.98, N = 12SE +/- 1.35, N = 1145.9245.8547.34MIN: 35.92 / MAX: 680.1MIN: 35.12 / MAX: 199.99MIN: 36.23 / MAX: 632.341. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: squeezenet_ssd1231020304050Min: 40.42 / Avg: 45.92 / Max: 55.45Min: 40.7 / Avg: 45.85 / Max: 51.54Min: 40.68 / Avg: 47.34 / Max: 52.961. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread

OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: regnety_400m1234080120160200SE +/- 5.19, N = 12SE +/- 3.47, N = 12SE +/- 3.25, N = 12193.29186.58184.95MIN: 146.8 / MAX: 9751.38MIN: 140.41 / MAX: 1358.11MIN: 150.03 / MAX: 1506.221. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread
OpenBenchmarking.orgms, Fewer Is BetterNCNN 20201218Target: CPU - Model: regnety_400m1234080120160200Min: 172.33 / Avg: 193.29 / Max: 226.5Min: 159.18 / Avg: 186.58 / Max: 203.27Min: 160.79 / Avg: 184.95 / Max: 196.541. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread