TNN

TNN is an open-source deep learning reasoning framework developed by Tencent.

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

Project Site

github.com

Test Created

24 September 2020

Last Updated

18 June 2021

Test Maintainer

Michael Larabel 

Test Type

System

Average Install Time

30 Seconds

Average Run Time

1 Minute, 11 Seconds

Test Dependencies

CMake + C/C++ Compiler Toolchain

Accolades

10k+ Downloads

Supported Platforms


Public Result Uploads *Reported Installs **Reported Test Completions **Test Profile Page Views ***OpenBenchmarking.orgEventsTNN Popularity Statisticspts/tnn2020.092020.112021.012021.032021.052021.072021.092021.112022.012022.032022.052022.072022.092022.112023.012023.032023.052023.072023.092023.112024.012024.032024.052024.072024.092K4K6K8K10K
* 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 updated weekly as of 3 October 2024.
MobileNet v224.9%SqueezeNet v1.124.2%DenseNet26.3%SqueezeNet v224.6%Model Option PopularityOpenBenchmarking.org

Revision History

pts/tnn-1.1.0   [View Source]   Fri, 18 Jun 2021 07:29:44 GMT
Update against TNN 0.3 upstream release.

pts/tnn-1.0.1   [View Source]   Mon, 11 Jan 2021 13:15:50 GMT
Update download mirror as the GitHub URL changed its checksums...

pts/tnn-1.0.0   [View Source]   Thu, 24 Sep 2020 18:33:29 GMT
Initial commit of Tencent TNN framework.

Suites Using This Test

Machine Learning

HPC - High Performance Computing


Performance Metrics

Analyze Test Configuration:

TNN 0.3

Target: CPU - Model: DenseNet

OpenBenchmarking.org metrics for this test profile configuration based on 1,025 public results since 18 June 2021 with the latest data as of 6 September 2024.

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)
94th
10
2062 +/- 92
92nd
7
2138 +/- 19
85th
61
2431 +/- 37
83rd
10
2460 +/- 283
81st
17
2531 +/- 24
Mid-Tier
75th
> 2603
75th
11
2609 +/- 56
75th
20
2618 +/- 58
75th
14
2634 +/- 36
72nd
14
2674 +/- 256
71st
18
2707 +/- 21
68th
12
2741 +/- 27
66th
4
2805 +/- 46
65th
18
2836 +/- 62
64th
4
2847 +/- 43
64th
6
2849 +/- 20
62nd
10
2868 +/- 3
58th
13
2922 +/- 113
56th
16
2984 +/- 94
56th
4
2987 +/- 39
55th
3
3009 +/- 2
51st
6
3058 +/- 1
Median
50th
3066
49th
5
3101 +/- 271
49th
4
3101 +/- 8
49th
3
3103 +/- 17
49th
7
3104 +/- 110
48th
4
3152 +/- 6
47th
6
3207 +/- 226
47th
7
3221 +/- 79
44th
6
3313 +/- 58
42nd
6
3407 +/- 31
42nd
3
3412 +/- 13
39th
3
3475 +/- 12
39th
8
3490 +/- 35
37th
3
3516 +/- 3
33rd
3
3674 +/- 4
33rd
3
3677 +/- 3
33rd
4
3687 +/- 1
32nd
6
3703 +/- 296
32nd
3
3707 +/- 10
32nd
3
3728 +/- 9
30th
3
3837 +/- 3
28th
3
3956 +/- 1
26th
3
4008 +/- 5
26th
3
4027 +/- 59
Low-Tier
25th
> 4035
25th
7
4049 +/- 109
25th
6
4081 +/- 12
25th
5
4085 +/- 80
23rd
6
4201 +/- 49
23rd
12
4210 +/- 219
23rd
3
4257 +/- 5
23rd
6
4258 +/- 90
22nd
4
4275 +/- 50
22nd
3
4290 +/- 7
21st
3
4319 +/- 19
21st
5
4365 +/- 483
18th
4
4467 +/- 26
18th
6
4503 +/- 67
17th
3
4523 +/- 5
17th
3
4565 +/- 5
16th
14
4612 +/- 494
15th
3
4738 +/- 8
15th
3
4743 +/- 12
15th
6
4864 +/- 445
15th
3
4873 +/- 44
14th
5
4897 +/- 26
14th
3
4999 +/- 7
13th
5
5058 +/- 84
13th
4
5128 +/- 3
12th
13
5201 +/- 268
11th
7
5275 +/- 160
9th
7
5520 +/- 27
8th
4
5818 +/- 65
8th
5
5886 +/- 88
7th
3
6188 +/- 41
7th
3
6358 +/- 949
5th
4
8347 +/- 9
4th
4
8548 +/- 10
2nd
12
20666 +/- 1638
OpenBenchmarking.orgDistribution Of Public Results - Target: CPU - Model: DenseNet1025 Results Range From 1418 To 129572 ms14183982654691101167414238168021936621930244942705829622321863475037314398784244245006475705013452698552625782660390629546551868082706467321075774783388090283466860308859491158937229628698850101414103978106542109106111670114234116798119362121926124490127054129618160320480640800

Based on OpenBenchmarking.org data, the selected test / test configuration (TNN 0.3 - Target: CPU - Model: DenseNet) has an average run-time of 15 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: DenseNetRun-Time20406080100Min: 5 / Avg: 14.22 / Max: 88

Based on public OpenBenchmarking.org results, the selected test / test configuration has an average standard deviation of 0.1%.

OpenBenchmarking.orgPercent, Fewer Is BetterAverage Deviation Between RunsTarget: CPU - Model: DenseNetDeviation246810Min: 0 / Avg: 0.09 / Max: 5

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 BaseTNN CPU Core ScalingTarget: CPU - Model: DenseNet246812162432640.71251.4252.13752.853.5625

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.
 
PUNPCKLQDQ MOVDQA MOVDQU PSRLDQ CVTSS2SD CVTSD2SS MOVAPD ADDSD UCOMISD SQRTSD MULSD DIVSD COMISD MOVD UNPCKLPD CVTPD2PS PSHUFD PMULUDQ PADDQ CVTDQ2PS CVTSI2SD CVTTSD2SI MOVUPD MAXSD XORPD SUBSD CVTTPS2DQ CVTPS2PD UNPCKHPD MULPD ANDPD CMPNLESD ANDNPD ORPD ADDPD SUBPD CVTDQ2PD CMPLEPD CVTTPD2DQ
Requires passing a supported compiler/build flag (verified with targets: sandybridge, skylake, tigerlake, cascadelake, sapphirerapids, alderlake, znver2, znver3).
Found on Intel processors since Sandy Bridge (2011).
Found on AMD processors since Bulldozer (2011).

 
VZEROUPPER VEXTRACTF128 VINSERTF128 VPERMILPS VPERM2F128 VBROADCASTSS VMASKMOVPS VBROADCASTSD
Requires passing a supported compiler/build flag (verified with targets: skylake, tigerlake, cascadelake, sapphirerapids, alderlake, znver2, znver3).
Found on Intel processors since Haswell (2013).
Found on AMD processors since Excavator (2016).

 
VEXTRACTI128 VPERMD VPBROADCASTQ VPERM2I128 VPBROADCASTD VPERMQ VINSERTI128 VPGATHERDD VPMASKMOVD VPBROADCASTW VPBROADCASTB
FMA (FMA)
Requires passing a supported compiler/build flag (verified with targets: skylake, tigerlake, cascadelake, sapphirerapids, alderlake, znver2, znver3).
Found on Intel processors since Haswell (2013).
Found on AMD processors since Bulldozer (2011).

 
VFMADD132PS VFNMADD132SS VFMADD231SS VFMADD231PS VFNMADD132PS VFMADD132SS VFMADD132SD VFMSUB231SD VFNMSUB231SD VFNMSUB132SD VFMADD132PD VFNMADD132PD VFMADD213PS VFMADD213SS VFNMADD231SD VFNMADD132SD VFMADD231PD VFMSUB231PS VFMSUB231SS VFNMADD231PD VFMADD231SD VFMSUB132SS VFMADD213SD
Advanced Vector Extensions 512 (AVX512)
Requires passing a supported compiler/build flag (verified with targets: cascadelake, sapphirerapids).
 
(ZMM REGISTER USE)
The test / benchmark does honor compiler flag changes.
Last automated analysis: 18 January 2022

This test profile binary relies on the shared libraries libTNN.so.0, libm.so.6, libc.so.6, libgomp.so.1.

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)
RISC-V 64-bit
riscv64
SiFive RISC-V
Loongson LoongArch 64-bit
loongarch64
Loongson-3A5000LL, Loongson-3C5000
ARMv7 32-bit
armv7l
ARMv7 Cortex-A72 4-Core, ARMv7 rev 1 4-Core
ARMv8 64-bit
aarch64
ARMv8 Cortex-A57 4-Core, ARMv8 Cortex-A72 16-Core, ARMv8 Cortex-A72 4-Core, ARMv8 Cortex-A76 4-Core, ARMv8 Cortex-A78E 6-Core, ARMv8 Neoverse-N1, ARMv8 Neoverse-N1 32-Core, ARMv8 Neoverse-N1 64-Core, ARMv8 Neoverse-V1, Ampere ARMv8 Neoverse-N1 256-Core, Ampere Altra ARMv8 Neoverse-N1 160-Core, Apple M1, Apple M2, HiSilicon TSV110, Phytium D2000, Rockchip ARMv8 Cortex-A76 4-Core, Snapdragon

Recent Test Results

OpenBenchmarking.org Results Compare

1 System - 4 Benchmark Results

Unknown - LOONGSON Dabieshan Loongson-TC542F0 - Loongson LLC Hyper Transport Bridge

Anolis OS 23.1 - 6.6.25-2.1.an23.loongarch64 - X Server 1.20.14

1 System - 4 Benchmark Results

HiSilicon TSV110 - Huawei TaiShan 200 - Huawei HiSilicon

Ubuntu 22.04 - 5.15.0-119-generic - 1.3.255

1 System - 334 Benchmark Results

Intel Core i3-12100 - ASRock H610M-HDV/M.2 R2.0 - Intel Device 7aa7

Ubuntu 20.04 - 5.15.0-89-generic - GNOME Shell 3.36.9

1 System - 275 Benchmark Results

Intel Core i5-12500 - ASUS PRIME Z690M-HZ - Intel Alder Lake-S PCH

Debian 12 - 6.1.0-21-amd64 - GCC 12.2.0

1 System - 4 Benchmark Results

Intel Core i7-12700H - Oracle VirtualBox v1.2 - Intel 440FX 82441FX PMC

Debian 12 - 6.1.0-18-amd64 - GNOME Shell 43.9

Find More Test Results