JPEG-XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library.

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

Project Site

jpegxl.info

Source Repository

github.com

Test Created

11 February 2021

Last Updated

1 March 2024

Test Maintainer

Michael Larabel 

Test Type

Processor

Test Dependencies

C/C++ Compiler Toolchain + JPEG Library + PNG Library + CMake + Curl

Accolades

40k+ Downloads

Supported Platforms


Public Result Uploads *Reported Installs **Reported Test Completions **Test Profile Page Views ***OpenBenchmarking.orgEventsJPEG-XL libjxl Popularity Statisticspts/jpegxl2021.022021.032021.042021.052021.062021.072021.082021.092021.102021.112021.122022.012022.022022.032022.042022.052022.062022.072022.082022.092022.102022.112022.122023.012023.022023.032023.042023.052023.062023.072023.082023.092023.102023.112023.122024.012024.022024.033K6K9K12K15K
* 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 25 March 2024.
8032.3%9032.7%10035.1%Quality Option PopularityOpenBenchmarking.org

Revision History

pts/jpegxl-1.6.0   [View Source]   Fri, 01 Mar 2024 14:54:05 GMT
Update against v0.10.1 upstream.

pts/jpegxl-1.5.0   [View Source]   Sat, 15 Oct 2022 05:53:54 GMT
Update against libjxl 0.7 upstream.

pts/jpegxl-1.4.0   [View Source]   Thu, 04 Nov 2021 14:09:36 GMT
Update against libjxl 0.6.1 upstream, also add Windows support now that there are official Windows binaries. Closes: https://github.com/phoronix-test-suite/test-profiles/issues/211

pts/jpegxl-1.3.0   [View Source]   Wed, 08 Sep 2021 09:08:42 GMT
Update against v0.5 upstream.

pts/jpegxl-1.2.0   [View Source]   Fri, 12 Mar 2021 07:11:13 GMT
Update against JPEG XL 0.3.3 upstream.

pts/jpegxl-1.1.1   [View Source]   Sat, 13 Feb 2021 10:58:33 GMT
Add curl as possible dependency due to the deps.sh using it to download dependencies.

pts/jpegxl-1.1.0   [View Source]   Fri, 12 Feb 2021 09:34:21 GMT
Add option of PNG and JPEG input files.

pts/jpegxl-1.0.1   [View Source]   Thu, 11 Feb 2021 16:14:17 GMT
Add Brotli as a system dependency.

pts/jpegxl-1.0.0   [View Source]   Thu, 11 Feb 2021 16:10:28 GMT
Initial commit of JPEG XL encode benchmark.

Suites Using This Test

Imaging

Creator Workloads


Performance Metrics

Analyze Test Configuration:

JPEG XL 0.3.3

Input: PNG - Encode Speed: 8

OpenBenchmarking.org metrics for this test profile configuration based on 99 public results since 12 March 2021 with the latest data as of 26 May 2021.

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
MP/s (Average)
96th
10
1.08 +/- 0.04
Mid-Tier
75th
< 0.86
64th
4
0.81 +/- 0.01
61st
3
0.76 +/- 0.01
Median
50th
0.68
46th
3
0.67 +/- 0.01
33rd
3
0.63 +/- 0.01
27th
3
0.61 +/- 0.01
Low-Tier
25th
< 0.61
20th
3
0.59 +/- 0.01
12th
4
0.52 +/- 0.01
8th
3
0.49 +/- 0.01
5th
3
0.48 +/- 0.01
OpenBenchmarking.orgDistribution Of Public Results - Input: PNG - Encode Speed: 899 Results Range From 0 To 2 MP/s0.1670.3340.5010.6680.8351.0021.1691.3361.5031.671.8372.004816243240

Based on OpenBenchmarking.org data, the selected test / test configuration (JPEG XL 0.3.3 - Input: PNG - Encode Speed: 8) has an average run-time of 19 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 BenchmarkInput: PNG - Encode Speed: 8Run-Time714212835Min: 13 / Avg: 18.58 / Max: 30

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 RunsInput: PNG - Encode Speed: 8Deviation246810Min: 0 / Avg: 0.15 / Max: 1

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.
 
CVTSI2SD MULSD SUBSD DIVSD CVTSS2SD ADDSD CVTSD2SS MOVD MOVDQU MOVDQA MOVAPD UCOMISD COMISD XORPD ANDPD MINSD MAXSD CVTTSD2SI SQRTSD CVTDQ2PS CVTTPS2DQ CVTPS2PD MULPD ADDPD ANDNPD CMPNLESD ORPD PSHUFD PMULUDQ CVTPS2DQ PADDQ PSLLDQ PUNPCKLQDQ PSRLDQ UNPCKLPD PUNPCKHQDQ PSUBQ SHUFPD MOVHPD MOVUPD
SSSE3 (SSSE3)
Used by default on supported hardware.
 
PALIGNR PSHUFB PABSD
SSE 4.2 (SSE4_2)
Used by default on supported hardware.
Found on Intel processors since at least 2010.
Found on AMD processors since Bulldozer (2011).

 
PCMPGTQ POPCNT
Used by default on supported hardware.
Found on Intel processors since Sandy Bridge (2011).
Found on AMD processors since Bulldozer (2011).

 
VBROADCASTSS VZEROUPPER VBROADCASTF128 VINSERTF128 VPERM2F128 VEXTRACTF128 VPERMILPS VPERMILPD
Used by default on supported hardware.
Found on Intel processors since Haswell (2013).
Found on AMD processors since Excavator (2016).

 
VPERM2I128 VPBROADCASTD VPERMPS VPERMQ VPBROADCASTQ VPERMD VPBLENDD VINSERTI128 VGATHERDPS VPGATHERDD VEXTRACTI128 VPGATHERDQ VPBROADCASTW VPBROADCASTB VPGATHERQD VPMASKMOVD VPSRLVQ VPGATHERQQ
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 VFMADD231PS VFMADD213SD VFMADD132SD VFMADD132SS VFNMADD132SD VFMADD213PS VFMSUB132PS VFMADD231SS VFMADD213SS VFMSUB231PS VFMSUB132SS VFNMADD132SS VFNMADD213PS VFNMADD213SS VFNMADD132PS VFMADD231PD VFMADD231SD VFMSUB213PS VFNMADD231PS VFNMSUB132PS VFNMSUB213PS VFMSUB231SD VFMSUB132SD VFNMADD213SD VFNMADD231SS VFNMADD231SD VFMSUB231SS
The test / benchmark does honor compiler flag changes.
Last automated analysis: 17 November 2022

This test profile binary relies on the shared libraries libjxl.so.0.7, libgif.so.7, libjpeg.so.8, libpng16.so.16, libm.so.6, libc.so.6, libbrotlidec.so.1, libbrotlienc.so.1, libz.so.1, libbrotlicommon.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)