oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the oneAPI initiative.

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

Project Site

github.com

Test Created

17 June 2020

Last Updated

20 December 2020

Test Maintainer

Michael Larabel

Test Type

Processor

Average Install Time

9 Minutes, 3 Seconds

Average Run Time

2 Minutes, 2 Seconds

Accolades

5k+ Downloads

Supported Platforms


Public Result UploadsReported Installs*Test Completions*OpenBenchmarking.orgEventsoneDNN Popularity Statisticspts/onednn2020.062020.072020.082020.092020.102020.112020.122021.014K8K12K16K20K
* 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.
Data current as of Mon, 18 Jan 2021 15:01:31 GMT.
Deconvolution Batch shapes_3d11.4%Recurrent Neural Network Training15.9%IP Shapes 1D11.5%Convolution Batch Shapes Auto11.5%Matrix Multiply Batch Shapes Transformer11.3%Recurrent Neural Network Inference15.7%IP Shapes 3D11.5%Deconvolution Batch shapes_1d11.4%Harness Option PopularityOpenBenchmarking.org
bf16bf16bf1614.1%u8s8f3240.0%f3245.9%Data Type Option PopularityOpenBenchmarking.org

Revision History

pts/onednn-1.6.1   [View Source]   Sun, 20 Dec 2020 09:58:16 GMT
This test profile builds and works fine on macOS so enable it (MacOSX).

pts/onednn-1.6.0   [View Source]   Wed, 09 Dec 2020 13:47:31 GMT
Update against oneDNN 2.0 upstream.

pts/onednn-1.5.0   [View Source]   Wed, 17 Jun 2020 16:26:39 GMT
Initial commit of oneDNN test profile based on Intel oneDNN 1.5, forked from existing mkl-dnn test profile that is named from MKL-DNN before it was renamed to DNNL and then oneDNN. So create new test profile to match Intel naming convention.

Suites Using This Test

Multi-Core

Machine Learning

Intel oneAPI

CPU Massive

Server CPU Tests

Creator Workloads

HPC - High Performance Computing


Performance Metrics

Analyze Test Configuration:

oneDNN 1.5

Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU

OpenBenchmarking.org metrics for this test profile configuration based on 92 public results since 20 June 2020 with the latest data as of 20 November 2020.

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
# Matching Public Results
ms (Average)
92nd
13
1.43 +/- 0.09
Mid-Tier
75th
> 1.67
64th
21
1.71 +/- 0.02
Median
50th
2.04
45th
6
2.13 +/- 0.24
Low-Tier
25th
> 3.50
22nd
5
3.57 +/- 0.18
18th
4
4.01 +/- 0.01
14th
5
5.39 +/- 0.03
6th
4
6.20 +/- 0.02
OpenBenchmarking.orgDistribution Of Public Results - Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU92 Results Range From 0 To 20 ms0246810121416182022241020304050

Based on OpenBenchmarking.org data, the selected test / test configuration (oneDNN 1.5 - Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU) has an average run-time of 2 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 BenchmarkHarness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPURun-Time246810Min: 1 / Avg: 1 / Max: 1

Based on public OpenBenchmarking.org results, the selected test / test configuration has an average standard deviation of 0.2%.

OpenBenchmarking.orgPercent, Fewer Is BetterAverage Deviation Between RunsHarness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPUDeviation246810Min: 0 / Avg: 0.22 / Max: 2

Recent Test Results

OpenBenchmarking.org Results Compare

1 System - 214 Benchmark Results

Intel Xeon Platinum 8275CL - Amazon EC2 c5.2xlarge - Intel 440FX 82441FX PMC

Ubuntu 20.04 - 5.4.0-1029-aws - GCC 9.3.0

1 System - 167 Benchmark Results

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

Debian testing - 5.8.0-2-amd64 - GNOME Shell 3.38.0

5 Systems - 158 Benchmark Results

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

Fedora 33 - 5.8.12-300.fc33.x86_64 - GNOME Shell 3.38.0

4 Systems - 158 Benchmark Results

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

Ubuntu 20.10 - 5.8.0-19-generic - GNOME Shell 3.38.0

3 Systems - 170 Benchmark Results

Intel Core i7-7900X - ASRock X299 Extreme4 - Intel Sky Lake-E DMI3 Registers

Ubuntu 19.04 - 5.0.0-38-generic - GNOME Shell 3.32.1

3 Systems - 155 Benchmark Results

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

Fedora 33 - 5.8.12-300.fc33.x86_64 - GNOME Shell 3.38.0

2 Systems - 182 Benchmark Results

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

Ubuntu 20.10 - 5.8.0-19-generic - GNOME Shell 3.38.0

1 System - 178 Benchmark Results

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

Ubuntu 20.10 - 5.8.0-19-generic - GNOME Shell 3.38.0

1 System - 20 Benchmark Results

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

Ubuntu 20.10 - 5.4.0-42-generic - GNOME Shell 3.36.4

1 System - 37 Benchmark Results

2 x Intel Xeon Gold 5220R - TYAN S7106 - Intel Sky Lake-E DMI3 Registers

Ubuntu 20.04 - 5.9.0-050900rc6-generic - GNOME Shell 3.36.4

1 System - 20 Benchmark Results

Most Popular Test Results

OpenBenchmarking.org Results Compare

6 Systems - 136 Benchmark Results

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

Debian testing - 5.7.0-1-amd64 - GNOME Shell 3.36.4

5 Systems - 158 Benchmark Results

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

openSUSE Tumbleweed 20200929 - 5.8.10-1-default - KDE Plasma 5.19.5

4 Systems - 158 Benchmark Results

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

Ubuntu 20.10 - 5.8.0-19-generic - GNOME Shell 3.38.0

3 Systems - 86 Benchmark Results

AMD Ryzen Threadripper 3960X 24-Core - MSI Creator TRX40 - AMD Starship

Ubuntu 20.04 - 5.4.0-39-generic - GNOME Shell 3.36.1

10 Systems - 130 Benchmark Results

2 x Intel Xeon Gold 5220R - GIGABYTE MD61-SC2-00 v01000100 - Intel Sky Lake-E DMI3 Registers

Ubuntu 18.04 - 5.3.0-40-generic - GNOME Shell 3.28.4

6 Systems - 158 Benchmark Results

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

Debian testing - 5.8.0-2-amd64 - GNOME Shell 3.38.0

1 System - 144 Benchmark Results

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

Clear Linux OS 33540 - 5.7.10-972.native - GNOME Shell 3.36.4

11 Systems - 138 Benchmark Results

2 x Intel Xeon Gold 6138 - GIGABYTE MD61-SC2-00 v01000100 - Intel Sky Lake-E DMI3 Registers

Ubuntu 18.04 - 5.3.0-40-generic - GNOME Shell 3.28.4

2 Systems - 86 Benchmark Results

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

Pop 20.04 - 5.4.0-7634-generic - GNOME Shell 3.36.3

10 Systems - 60 Benchmark Results

2 x Intel Xeon Gold 6258R - GIGABYTE MD61-SC2-00 v01000100 - Intel Sky Lake-E DMI3 Registers

Ubuntu 18.04 - 5.3.0-40-generic - GNOME Shell 3.28.4

1 System - 194 Benchmark Results

4 x Intel Xeon Gold 5218R - Intel 440BX - Intel 440BX

CentOS Linux 8 - 4.18.0-193.6.3.el8_2.x86_64 - GCC 8.3.1 20191121

Find More Test Results

OpenBenchmarking.org Community User Comments

Post A Comment