NCNN

NCNN is a high performance neural network inference framework optimized for mobile and other platforms developed by Tencent.

To run this test with the Phoronix Test Suite, the basic command is: phoronix-test-suite benchmark ncnn.

Project Site

github.com

Source Repository

github.com

Test Created

18 September 2020

Last Updated

13 August 2022

Test Maintainer

Michael Larabel 

Test Type

System

Average Install Time

1 Minute, 11 Seconds

Average Run Time

20 Minutes, 48 Seconds

Test Dependencies

CMake + C/C++ Compiler Toolchain + Vulkan

Accolades

20k+ Downloads

Supported Platforms


Public Result Uploads *Reported Installs **Reported Test Completions **Test Profile Page Views ***OpenBenchmarking.orgEventsNCNN Popularity Statisticspts/ncnn2020.092020.102020.112020.122021.012021.022021.032021.042021.052021.062021.072021.082021.092021.102021.112021.122022.012022.022022.032022.042022.052022.062022.072022.082022.092022.102022.115K10K15K20K25K
* Uploading of benchmark result data to OpenBenchmarking.org is always optional (opt-in) via the Phoronix Test Suite for users wishing to share their results publicly.
** Data based on those opting to upload their test results to OpenBenchmarking.org and users enabling the opt-in anonymous statistics reporting while running benchmarks from an Internet-connected platform.
*** Test profile page view reporting began March 2021.
Data current as of 26 November 2022.
Vulkan GPU23.1%CPU76.9%Target Option PopularityOpenBenchmarking.org
efficientnet-b05.6%regnety_400m5.6%alexnet6.1%googlenet5.6%mobilenet-v35.6%blazeface5.6%vgg166.3%mobilenet6.3%FastestDet5.4%squeezenet_ssd6.1%yolov4-tiny6.1%resnet506.3%mobilenet-v25.7%mnasnet5.6%vision_transformer6.3%resnet186.2%shufflenet-v25.6%Model Option PopularityOpenBenchmarking.org

Revision History

pts/ncnn-1.4.0   [View Source]   Sat, 13 Aug 2022 09:59:41 GMT
Update against NCNN 20220729 upstream.

pts/ncnn-1.3.0   [View Source]   Tue, 27 Jul 2021 16:34:42 GMT
Update against NCNN 2021-07-20 upstream, fix possible Vulkan build issue by including glslang source tree.

pts/ncnn-1.2.0   [View Source]   Fri, 18 Jun 2021 08:11:33 GMT
Update against NCNN 20210525 release.

pts/ncnn-1.1.0   [View Source]   Fri, 18 Dec 2020 08:06:41 GMT
Update against new upstream NCNN 20201218.

pts/ncnn-1.0.3   [View Source]   Fri, 25 Sep 2020 06:36:39 GMT
Drop int8 tests per https://github.com/phoronix-test-suite/test-profiles/pull/167

pts/ncnn-1.0.2   [View Source]   Thu, 24 Sep 2020 12:52:47 GMT
Expose Vulkan GPU support.

pts/ncnn-1.0.1   [View Source]   Fri, 18 Sep 2020 12:28:10 GMT
Increase the run count.

pts/ncnn-1.0.0   [View Source]   Fri, 18 Sep 2020 11:58:15 GMT
Initial commit of Tencent NCNN.

Suites Using This Test

Machine Learning

HPC - High Performance Computing

NVIDIA GPU Compute

Vulkan Compute


Performance Metrics

Analyze Test Configuration:

NCNN 20220729

Target: CPU - Model: mobilenet

OpenBenchmarking.org metrics for this test profile configuration based on 277 public results since 13 August 2022 with the latest data as of 19 November 2022.

Below is an overview of the generalized performance for components where there is sufficient statistically significant data based upon user-uploaded results. It is important to keep in mind particularly in the Linux/open-source space there can be vastly different OS configurations, with this overview intended to offer just general guidance as to the performance expectations.

Component
Percentile Rank
# Compatible Public Results
ms (Average)
Mid-Tier
75th
> 10
58th
7
12
Median
50th
13
33rd
4
16 +/- 1
31st
3
17 +/- 1
Low-Tier
25th
> 20
7th
3
33 +/- 1
5th
3
36 +/- 1
OpenBenchmarking.orgDistribution Of Public Results - Target: CPU - Model: mobilenet277 Results Range From 6 To 188 ms6193245587184971101231361491621751884080120160200

Based on OpenBenchmarking.org data, the selected test / test configuration (NCNN 20220729 - Target: CPU - Model: mobilenet) has an average run-time of 12 minutes. By default this test profile is set to run at least 3 times but may increase if the standard deviation exceeds pre-defined defaults or other calculations deem additional runs necessary for greater statistical accuracy of the result.

OpenBenchmarking.orgMinutesTime Required To Complete BenchmarkTarget: CPU - Model: mobilenetRun-Time1224364860Min: 2 / Avg: 11.84 / Max: 60

Based on public OpenBenchmarking.org results, the selected test / test configuration has an average standard deviation of 0.5%.

OpenBenchmarking.orgPercent, Fewer Is BetterAverage Deviation Between RunsTarget: CPU - Model: mobilenetDeviation3691215Min: 0 / Avg: 0.48 / Max: 9

Does It Scale Well With Increasing Cores?

No, based on the automated analysis of the collected public benchmark data, this test / test settings does not generally scale well with increasing CPU core counts. Data based on publicly available results for this test / test settings, separated by vendor, result divided by the reference CPU clock speed, grouped by matching physical CPU core count, and normalized against the smallest core count tested from each vendor for each CPU having a sufficient number of test samples and statistically significant data.

IntelAMDOpenBenchmarking.orgRelative Core Scaling To BaseNCNN CPU Core ScalingTarget: CPU - Model: mobilenet4681216320.71461.42922.14382.85843.573

Notable Instruction Set Usage

Notable instruction set extensions supported by this test, based on an automatic analysis by the Phoronix Test Suite / OpenBenchmarking.org analytics engine.

Instruction Set
Support
Instructions Detected
SSE2 (SSE2)
Used by default on supported hardware.
 
MOVD PSHUFD PSHUFLW MOVDQU MOVDQA PUNPCKLQDQ SUBSD MOVAPD MINSD MAXSD ADDSD CVTSI2SD DIVSD CVTSS2SD PADDQ PSUBQ SHUFPD MULSD CVTTSD2SI MULPD PSRLDQ CVTSD2SS SQRTSD CVTTPS2DQ CVTDQ2PS PUNPCKHQDQ PMULUDQ COMISD ANDPD UCOMISD MOVUPD MOVHPD UNPCKHPD CMPNLESD ANDNPD ORPD XORPD SUBPD
Used by default on supported hardware.
Found on Intel processors since Sandy Bridge (2011).
Found on AMD processors since Bulldozer (2011).

 
VZEROUPPER VBROADCASTSS VINSERTF128 VEXTRACTF128 VPERM2F128 VPERMILPS VMASKMOVPS VBROADCASTSD
Used by default on supported hardware.
Found on Intel processors since Haswell (2013).
Found on AMD processors since Excavator (2016).

 
VPBROADCASTD VPERMD VEXTRACTI128 VINSERTI128 VPERMQ VPERM2I128 VPBROADCASTW VPBROADCASTQ VPBROADCASTB VGATHERDPS VPGATHERQQ VPSRLVQ VPBLENDD VPGATHERQD VPERMPD
Advanced Vector Extensions 512 (AVX512)
Used by default on supported hardware.
 
(ZMM REGISTER USE)
FMA (FMA)
Used by default on supported hardware.
Found on Intel processors since Haswell (2013).
Found on AMD processors since Bulldozer (2011).

 
VFMADD132PS VFMADD132SS VFMADD231SS VFNMADD231PS VFMADD231PS VFMADD213PS VFNMADD132PS VFMSUB132PS VFNMADD231SS VFMADD213SS VFNMADD132SS VFMSUB231SS VFMSUB132SS VFMSUB213SS VFNMADD213SS VFMADD132PD VFMADD132SD VFNMSUB132SS VFNMSUB231SS VFMADD231SD VFNMADD132SD VFMSUB231SD VFNMADD213SD
AVX Vector Neural Network Instructions (AVX-VNNI)
Requires passing a supported compiler/build flag (verified with targets: tigerlake, cascadelake).
 
VPDPWSSD
The test / benchmark does honor compiler flag changes.
Last automated analysis: 13 August 2022

This test profile binary relies on the shared libraries libgomp.so.1, libvulkan.so.1, libm.so.6, libmvec.so.1, libc.so.6.

Tested CPU Architectures

This benchmark has been successfully tested on the below mentioned architectures. The CPU architectures listed is where successful OpenBenchmarking.org result uploads occurred, namely for helping to determine if a given test is compatible with various alternative CPU architectures.

CPU Architecture
Kernel Identifier
Verified On
Intel / AMD x86 64-bit
x86_64
(Many Processors)
ARMv8 64-bit
arm64
Apple M1
ARMv8 64-bit
aarch64
ARMv8 Cortex-A72 4-Core, Apple M1, Apple M2

Recent Test Results

OpenBenchmarking.org Results Compare

2 Systems - 334 Benchmark Results

2 x AMD EPYC 7601 32-Core - Dell 02MJ3T - AMD 17h

Ubuntu 22.04 - 5.15.0-40-generic - GNOME Shell 42.2

2 Systems - 334 Benchmark Results

2 x AMD EPYC 7601 32-Core - Dell 02MJ3T - AMD 17h

Ubuntu 22.04 - 5.15.0-40-generic - GNOME Shell 42.2

1 System - 78 Benchmark Results

Intel Core i5-4250U - Apple MacBook Air - Apple SSD

macOS 10.15.7 - 19.6.0 - Clang 12.0.0 + GCC 12.0.0

2 Systems - 159 Benchmark Results

Intel Xeon E5-2609 v4 - MSI X99A RAIDER - Intel Xeon E7 v4

Ubuntu 20.04 - 5.9.0-050900rc6daily20200926-generic - GNOME Shell 3.36.2

1 System - 149 Benchmark Results

Apple M1 - Apple MacBook Pro - 16GB

macOS 11.7 - 20.6.0 - GCC 13.0.0 + Clang 13.0.0

1 System - 78 Benchmark Results

Intel Core i5-4250U - Apple MacBook Air - Apple SSD

macOS 10.15.7 - 19.6.0 - Clang 12.0.0 + GCC 12.0.0

1 System - 78 Benchmark Results

Apple M1 - Apple MacBook Pro - 16GB

macOS 11.7 - 20.6.0 - GCC 13.0.0 + Clang 13.0.0

1 System - 147 Benchmark Results

Apple M1 - Apple Mac mini - 8GB

macOS 12.0.1 - 21.1.0 - GCC 13.0.0 + Clang 13.0.0 + Xcode 13.1

1 System - 704 Benchmark Results

Intel Core i9-11900K - ASUS ROG MAXIMUS XIII HERO - Intel Tiger Lake-H

Ubuntu 22.10 - 5.19.0-21-generic - GNOME Shell 43.0

1 System - 147 Benchmark Results

Apple M1 - Apple Mac mini - 8GB

macOS 12.0.1 - 21.1.0 - GCC 13.0.0 + Clang 13.0.0 + Xcode 13.1

3 Systems - 213 Benchmark Results

AMD Ryzen 7 4800U - ASRock 4X4-4000 - AMD Renoir

Ubuntu 22.04 - 5.19.0-rc6-phx-retbleed - GNOME Shell 42.2

1 System - 100 Benchmark Results

AMD Ryzen 5 7600X 6-Core - ASUS ROG CROSSHAIR X670E HERO - AMD Device 14d8

Ubuntu 22.04 - 6.0.0-060000rc1daily20220820-generic - GNOME Shell 42.2

3 Systems - 357 Benchmark Results

AMD Ryzen 7 7700X 8-Core - ASUS ROG CROSSHAIR X670E HERO - AMD Device 14d8

Ubuntu 22.04 - 6.0.0-060000rc1daily20220820-generic - GNOME Shell 42.2

3 Systems - 357 Benchmark Results

Most Popular Test Results

OpenBenchmarking.org Results Compare

13 Systems - 333 Benchmark Results

AMD Ryzen 7 5800X 8-Core - ASUS ROG CROSSHAIR VIII HERO - AMD Starship

Ubuntu 22.04 - 6.0.0-060000rc1daily20220820-generic - GNOME Shell 42.2

3 Systems - 63 Benchmark Results

AMD Ryzen Threadripper 2990WX 32-Core - ASUS ROG ZENITH EXTREME - AMD 17h

Ubuntu 22.04 - 5.19.0-051900rc7-generic - GNOME Shell 42.2

3 Systems - 63 Benchmark Results

Intel Xeon E3-1275 v6 - ASUS P10S-M WS - Intel Xeon E3-1200 v6

Ubuntu 20.04 - 5.9.0-050900rc8daily20201007-generic - X Server 1.20.8

3 Systems - 78 Benchmark Results

Intel Core i7-5960X - Gigabyte X99-UD4-CF - Intel Xeon E7 v3

Debian 11 - 5.10.0-10-amd64 - 1.0.2

4 Systems - 63 Benchmark Results

Intel Xeon Silver 4216 - TYAN S7100AG2NR - Intel Sky Lake-E DMI3 Registers

Debian 11 - 5.10.0-10-amd64 - X Server

3 Systems - 311 Benchmark Results

Intel Core i5-12600K - ASUS PRIME Z690-P WIFI D4 - Intel Device 7aa7

Ubuntu 22.04 - 5.19.0-051900rc6daily20220716-generic - GNOME Shell 42.1

16 Systems - 333 Benchmark Results

AMD Ryzen 7 5700G - ASUS ROG CROSSHAIR VIII HERO - AMD Renoir

Ubuntu 22.04 - 6.0.0-060000rc1daily20220820-generic - GNOME Shell 42.2

4 Systems - 63 Benchmark Results

AMD EPYC 7F32 8-Core - ASRockRack EPYCD8 - AMD Starship

Debian 11 - 5.10.0-10-amd64 - GNOME Shell 3.38.6

3 Systems - 307 Benchmark Results

AMD Ryzen 9 3900X 12-Core - ASUS TUF GAMING X570-PLUS - AMD Starship

Ubuntu 20.04 - 5.11.0-rc1-phx - GNOME Shell 3.36.4

3 Systems - 32 Benchmark Results

Intel Core i9-10980XE - ASRock X299 Steel Legend - Intel Sky Lake-E DMI3 Registers

Ubuntu 22.04 - 5.19.0-051900rc7-generic - GNOME Shell 42.2

3 Systems - 46 Benchmark Results

AMD Ryzen 5 5500U - NB01 NL5xNU - AMD Renoir

Ubuntu 22.04 - 5.18.8-051808-generic - GNOME Shell 42.2

4 Systems - 211 Benchmark Results

AMD Ryzen 9 3900XT 12-Core - MSI MEG X570 GODLIKE - AMD Starship

Ubuntu 22.04 - 5.15.0-40-generic - GNOME Shell 42.2

Find More Test Results