Performance Metrics

As of release 24.11, for each release, USD generates performance metrics using a specific set of assets and specific hardware and software configurations. This page describes what metrics are collected, what hardware and software configurations are used, the actual metrics results, and how to generate the metrics locally.

What We Measure

For a given asset, our performance script captures the following metrics by default (items in bold are reported on this page):

• Time to load and configure USD plugins for usdview

• Time to open the stage

• Time to reset the usdview prim browser

• Time to initialize the usdview UI

• Time to render first image (in usdview)

• Time to shutdown Hydra

• Time to close the stage

• Time to tear down the usdview UI

• Total time to start and quit usdview

• Time to traverse the prims in the stage

We run 100 iterations for each asset, and capture the minimum and maximum times for that set of iterations. We also calculate the mean time across the 100 iterations.

For each asset, we first warm the filesystem cache by loading the asset in usdview, to ensure we’re not including cache performance issues in our metrics.

All assets used to measure performance are assumed to be available locally. Time to download assets is not measured or included as part of the gathered performance metrics.

What Environment Is Used

This section describes the computing environment used to generate the published performance metrics. The following operating systems and hardware are currently used.

Note

Machine specifications are subject to change. If specifications do change for a give release, historical performance measurements will be run to backfill any outdated data.

Linux

• OS: CentOS Linux 7

• CPU: AMD EPYC 7763 64-Core Processor, 2450 Mhz

• CPU Utilization: 31 Core(s), 31 Logical Processor(s) (no hyperthreading)

• RAM: 117GB

• GPU: NVIDIA RTXA6000-24Q

macOS

• OS: macOS 14.3

• CPU: Apple M2 Ultra (20 Core)

• RAM: 192GB

• GPU: Apple M2 Ultra GPU (76 Core)

Windows

• OS: Microsoft Windows 11 Enterprise

• CPU: AMD EPYC 7763 64-Core Processor, 2450 Mhz

• CPU Utilization: 31 Core(s), 31 Logical Processor(s) (no hyperthreading)

• RAM: 128GB

• GPU: NVIDIA RTXA6000-24Q

USD Build

For each of the operating systems and hardware platforms listed previously, we build USD with the same build configuration. We use a stock invocation of build_usd.py with the default options (release build, Python components, imaging and USD imaging components, usdview, etc). Note that this build uses the default system memory allocator, and does not use an alternate allocator (as described in Use an allocator optimized for multithreading).

Metrics

Metrics are all measured in seconds.

Performance Graphs Per Platform

The following graphs show the time (in seconds) to open and close usdview for each asset. Graphs are provided for Linux, macOS, and Windows platforms (as described in What Environment Is Used). Performance data from the four most recent releases is reported on a rolling basis.

Note

A full historical rerun of all metrics on this page was performed for the 25.11 release because of a change in “create first image” timing methodology.

Interpret the metrics on this page with caution. Variation may arise from a variety of sources, including run-to-run inconsistency in how the GL driver schedules its flushes against our requests. Investigation on these is ongoing.

Standard Shader Ball

This asset is designed to be a comprehensive test of a broad array of material properties in a single render. Geometry is expressed using USD, materials are defined using MaterialX, texture maps are provided in OpenEXR format and encoded using the Academy Color Encoding System ACEScg color space.

In our performance sweep we use the mtlx_bubblegum material. Note the image below may not exactly match the image generated during the performance run; the primary intention is not necessarily to exercise specific renderers.

The shader ball asset can be downloaded here.

Linux Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.115236 max: 0.138383 mean: 0.12241002	min: 0.107972 max: 0.121466 mean: 0.11391579	min: 0.107262 max: 0.121741 mean: 0.11261217	min: 0.130829 max: 0.150694 mean: 0.13686351
Render first image	min: 1.014073 max: 1.174135 mean: 1.03070467	min: 1.158713 max: 1.244631 mean: 1.18729007	min: 1.158948 max: 1.267224 mean: 1.19264357	min: 1.238056 max: 6.023667 mean: 1.30996932
Close stage	min: 0.000419 max: 0.00088 mean: 0.00063121	min: 0.000403 max: 0.00098 mean: 0.00063492	min: 0.00041 max: 0.000872 mean: 0.00063941	min: 0.000299 max: 0.000758 mean: 0.00053994
Shut down Hydra	min: 0.015888 max: 0.020628 mean: 0.01816711	min: 0.019347 max: 0.025039 mean: 0.0214453	min: 0.019558 max: 0.025509 mean: 0.02220578	min: 0.01957 max: 0.026398 mean: 0.02231206

macOS Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.09158 max: 0.09794 mean: 0.09362494	min: 0.089313 max: 0.099611 mean: 0.0922419	min: 0.08957 max: 0.094074 mean: 0.09161377	min: 0.096033 max: 0.106589 mean: 0.10036488
Render first image	min: 0.707172 max: 0.743281 mean: 0.72663081	min: 0.732215 max: 1.182054 mean: 0.75341856	min: 0.734351 max: 0.771068 mean: 0.75320206	min: 1.088926 max: 1.434863 mean: 1.11545414
Close stage	min: 0.000227 max: 0.000438 mean: 0.00027077	min: 0.000228 max: 0.00072 mean: 0.0002692	min: 0.000229 max: 0.000765 mean: 0.00027705	min: 0.000119 max: 0.000491 mean: 0.00016517
Shut down Hydra	min: 0.010882 max: 0.018872 mean: 0.01490757	min: 0.012043 max: 0.020186 mean: 0.01560325	min: 0.012992 max: 0.021094 mean: 0.01651058	min: 0.012247 max: 0.020553 mean: 0.01637388

Windows Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.25587 max: 0.305149 mean: 0.26487604	min: 0.261208 max: 0.285934 mean: 0.27106556	min: 0.307208 max: 0.330622 mean: 0.31348188	min: 0.298701 max: 0.315642 mean: 0.30742671
Render first image	min: 1.992974 max: 2.180966 mean: 2.07286652	min: 2.147584 max: 2.276122 mean: 2.20730552	min: 2.200855 max: 2.395714 mean: 2.29471476	min: 2.184459 max: 8.745092 mean: 2.3044740200000002
Close stage	min: 0.001053 max: 0.001517 mean: 0.0011917100000000001	min: 0.001098 max: 0.003072 mean: 0.00161325	min: 0.001129 max: 0.002772 mean: 0.00170384	min: 0.000316 max: 0.001709 mean: 0.00068915
Shut down Hydra	min: 0.054428 max: 0.061991 mean: 0.05740973	min: 0.060869 max: 0.070369 mean: 0.06414039	min: 0.062921 max: 0.072712 mean: 0.06641849	min: 0.060228 max: 0.069912 mean: 0.06285791

Kitchen Set

This asset provides a complex kitchen scene.

The Kitchen Set asset can be downloaded here.

Linux Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.073168 max: 0.145009 mean: 0.09162708	min: 0.074726 max: 0.144169 mean: 0.09645783	min: 0.068395 max: 0.130942 mean: 0.08763188	min: 0.067288 max: 0.141716 mean: 0.08831019
Render first image	min: 0.226054 max: 0.265641 mean: 0.24998339	min: 0.22368 max: 0.278274 mean: 0.25676237	min: 0.232171 max: 0.277937 mean: 0.25756001	min: 0.269004 max: 0.370133 mean: 0.3244681
Close stage	min: 0.005919 max: 0.009146 mean: 0.00745146	min: 0.006153 max: 0.00939 mean: 0.00767703	min: 0.006141 max: 0.011938 mean: 0.00778345	min: 0.006306 max: 0.00948 mean: 0.00794424
Shut down Hydra	min: 0.009327 max: 0.01277 mean: 0.01035587	min: 0.019028 max: 0.025645 mean: 0.02086288	min: 0.022091 max: 0.028661 mean: 0.02401296	min: 0.01889 max: 0.027431 mean: 0.02081045

macOS Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.066191 max: 0.092933 mean: 0.07308952	min: 0.065663 max: 0.090223 mean: 0.07125644	min: 0.063966 max: 0.080906 mean: 0.0698448	min: 0.061154 max: 0.084453 mean: 0.0677701
Render first image	min: 0.265819 max: 0.27916 mean: 0.27261008999999997	min: 0.266764 max: 0.283842 mean: 0.27376717	min: 0.269227 max: 0.290286 mean: 0.27491572999999997	min: 0.259431 max: 0.27281 mean: 0.26522798000000003
Close stage	min: 0.002104 max: 0.003745 mean: 0.00246378	min: 0.002111 max: 0.003524 mean: 0.00255995	min: 0.002026 max: 0.003375 mean: 0.00240484	min: 0.002038 max: 0.003781 mean: 0.00244283
Shut down Hydra	min: 0.008501 max: 0.018295 mean: 0.01436839	min: 0.014569 max: 0.025991 mean: 0.01954361	min: 0.016196 max: 0.029243 mean: 0.02228483	min: 0.017222 max: 0.02856 mean: 0.023692190000000002

Windows Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.107719 max: 0.139995 mean: 0.11558934	min: 0.113508 max: 0.133956 mean: 0.12159507	min: 0.111981 max: 0.137971 mean: 0.12011667	min: 0.105351 max: 0.120111 mean: 0.11219626
Render first image	min: 0.361473 max: 0.394963 mean: 0.37519973	min: 0.369105 max: 0.398015 mean: 0.38108189	min: 0.370883 max: 0.410614 mean: 0.38819254	min: 0.387114 max: 0.430436 mean: 0.40021584
Close stage	min: 0.017472 max: 0.025552 mean: 0.02119207	min: 0.017828 max: 0.024302 mean: 0.02113338	min: 0.018721 max: 0.026777 mean: 0.02178006	min: 0.01869 max: 0.028976 mean: 0.02328194
Shut down Hydra	min: 0.022411 max: 0.028659 mean: 0.0245225	min: 0.034458 max: 0.048988 mean: 0.03803212	min: 0.040022 max: 0.047903 mean: 0.04345859	min: 0.045518 max: 0.05392 mean: 0.049226980000000004

ALab

ALab is a full production scene created by Animal Logic and contains over 300 assets, complete with high-quality textures and two characters with looping animation in shot context. Supplied as four separate downloads: the full production scene, high-quality textures, shot cameras matching the ALab trailer, and baked procedural fur and fabric for the animated characters.

The metrics have been measured with the base asset merged with the additional “techvars” info.

The ALab asset can be downloaded here.

Linux Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.371357 max: 0.548271 mean: 0.45959872	min: 0.548911 max: 0.678412 mean: 0.6064023	min: 0.370885 max: 0.607741 mean: 0.49451881	min: 0.433278 max: 0.610651 mean: 0.49882659
Render first image	min: 5.28362 max: 6.896981 mean: 6.13319128	min: 5.766128 max: 6.884635 mean: 6.19572927	min: 5.739026 max: 6.745488 mean: 6.14509327	min: 6.466451 max: 7.045572 mean: 6.64202786
Close stage	min: 0.077038 max: 0.092187 mean: 0.08111201	min: 0.073073 max: 0.088197 mean: 0.07835102000000001	min: 0.072466 max: 0.090774 mean: 0.07892472	min: 0.071978 max: 0.084795 mean: 0.07726876
Shut down Hydra	min: 0.055039 max: 0.079683 mean: 0.06659816	min: 0.102756 max: 0.125685 mean: 0.11473493	min: 0.112753 max: 0.145573 mean: 0.12437892	min: 0.138889 max: 0.1682 mean: 0.15461057

macOS Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.424971 max: 0.486238 mean: 0.4425593	min: 0.448676 max: 0.696681 mean: 0.48802467	min: 0.404042 max: 0.525631 mean: 0.43365319	min: 0.387289 max: 0.476831 mean: 0.41920031
Render first image	min: 2.64057 max: 3.143005 mean: 2.89778935	min: 2.657646 max: 3.089222 mean: 2.88767524	min: 2.692878 max: 3.149042 mean: 2.90695481	min: 3.374587 max: 3.796386 mean: 3.59016131
Close stage	min: 0.034361 max: 0.049179 mean: 0.03895347	min: 0.033888 max: 0.04874 mean: 0.03857322	min: 0.032976 max: 0.045111 mean: 0.03838184	min: 0.032904 max: 0.046616 mean: 0.03683789
Shut down Hydra	min: 0.095747 max: 0.118626 mean: 0.10839144	min: 0.132362 max: 0.15432 mean: 0.14241809	min: 0.125249 max: 0.160093 mean: 0.14434834	min: 0.142813 max: 0.177915 mean: 0.15281977

Windows Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.380082 max: 0.426562 mean: 0.39697795	min: 0.462555 max: 0.49612 mean: 0.47833338	min: 0.388397 max: 0.501158 mean: 0.40847451	min: 0.388077 max: 0.416403 mean: 0.4029927
Render first image	min: 7.614644 max: 8.894503 mean: 7.89112705	min: 7.725252 max: 8.814743 mean: 7.86926245	min: 7.661506 max: 8.938386 mean: 7.95734312	min: 7.782884 max: 8.019821 mean: 7.89220199
Close stage	min: 0.289758 max: 0.374639 mean: 0.33625491	min: 0.299041 max: 0.361021 mean: 0.33138915	min: 0.309628 max: 0.373266 mean: 0.33488903000000003	min: 0.317541 max: 0.405188 mean: 0.35938741
Shut down Hydra	min: 0.210191 max: 0.448725 mean: 0.29269323	min: 0.268075 max: 0.511238 mean: 0.35748124	min: 0.275811 max: 0.522579 mean: 0.36632579	min: 0.329635 max: 0.572838 mean: 0.44994447

Moore Lane

4004 Moore Lane is a fully composed, high-quality scene for the purpose of testing various visual computing issues. The house itself was wrapped around a number of typical problem areas for light transport and noise sampling. This includes things like thin openings in exterior walls, recessed area light sources, deeply shadowed corners, and high-frequency details. The exterior landscape surrounding the house consisted of a relatively simple ecosystem of instanced plants which could provide some additional levels of complexity. In addition to the geometry itself being designed to exacerbate some typical issues, the USD structure itself was created for several layers of testing.

The metrics have been measured using the contained MooreLane_ASWF_0623.usda file.

The Moore Lane asset can be downloaded here.

Linux Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.059526 max: 0.107525 mean: 0.07169211	min: 0.057381 max: 0.104375 mean: 0.06832833	min: 0.057288 max: 0.107432 mean: 0.06790111	min: 0.055827 max: 0.112216 mean: 0.06557084
Render first image	min: 10.264518 max: 10.814635 mean: 10.47099948	min: 10.309351 max: 10.822443 mean: 10.51112802	min: 10.277389 max: 10.730336 mean: 10.50606514	min: 8.860546 max: 9.682169 mean: 9.46078172
Close stage	min: 0.008411 max: 0.010604 mean: 0.00929994	min: 0.008767 max: 0.010335 mean: 0.00938169	min: 0.00855 max: 0.009954 mean: 0.00924602	min: 0.008235 max: 0.011516 mean: 0.00891237
Shut down Hydra	min: 0.040028 max: 0.181531 mean: 0.06746142	min: 0.045026 max: 0.16766 mean: 0.07075642	min: 0.047149 max: 0.145957 mean: 0.06069932	min: 0.047152 max: 0.201449 mean: 0.09379694

macOS Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.089092 max: 0.093273 mean: 0.09061327	min: 0.08339 max: 0.08886 mean: 0.08464025	min: 0.082767 max: 0.086665 mean: 0.08393583	min: 0.0779 max: 0.083264 mean: 0.07919203
Render first image	min: 4.863602 max: 5.776998 mean: 5.27400174	min: 4.842576 max: 5.902118 mean: 5.2332561	min: 4.836536 max: 5.79493 mean: 5.2620495	min: 4.928251 max: 5.818517 mean: 5.27123588
Close stage	min: 0.003309 max: 0.012977 mean: 0.00439985	min: 0.003316 max: 0.010872 mean: 0.00414966	min: 0.003405 max: 0.008938 mean: 0.0039708	min: 0.003165 max: 0.008126 mean: 0.0034966899999999998
Shut down Hydra	min: 0.09596 max: 0.375403 mean: 0.30230101	min: 0.094684 max: 0.372043 mean: 0.28783911	min: 0.115016 max: 0.220803 mean: 0.12306854	min: 0.121073 max: 0.137621 mean: 0.12819004

Windows Metrics (min/max/mean in seconds)

Metric	25.02	25.05	25.08	25.11
Open stage	min: 0.143463 max: 0.167457 mean: 0.15291026	min: 0.138429 max: 0.153354 mean: 0.14575919	min: 0.143945 max: 0.169911 mean: 0.1492643	min: 0.135837 max: 0.148618 mean: 0.14251247
Render first image	min: 9.360458 max: 11.134516 mean: 10.02221568	min: 9.49156 max: 11.177957 mean: 10.24690578	min: 8.738549 max: 10.169592 mean: 9.10423748	min: 8.113206 max: 9.48835 mean: 8.81574719
Close stage	min: 0.094321 max: 0.117807 mean: 0.10088475	min: 0.095224 max: 0.119185 mean: 0.10143141	min: 0.094702 max: 0.116359 mean: 0.10147604	min: 0.094986 max: 0.124853 mean: 0.10156746999999999
Shut down Hydra	min: 0.310136 max: 0.4116 mean: 0.36513765	min: 0.311263 max: 0.40723 mean: 0.36185346	min: 0.31048 max: 0.415117 mean: 0.36670874999999997	min: 0.321071 max: 0.470964 mean: 0.38557374

Running Performance Metrics Locally

We encourage developers to run the USD performance metrics to measure performance impacts of OpenUSD code contributions. Performance metrics can also be run to validate local runtime environments and hardware configurations.

Performance metrics are generating using the usdmeasureperformance.py script found in pxr/extras/performance. See the usdmeasureperformance tool docs for more information on the different parameters available.

usdmeasureperformance.py uses usdview and testusdview, so you will need to make sure those are in your current path, or aliased properly.

For gathering the metrics published on this page, the following parameters are used (for each asset):

python usdmeasureperformance.py <asset.usda> -i 100 -a min -o <metrics output filename.yaml>

Optionally, –tracedir <dir> will output trace information to dir that may aid in performance debugging.

Adding Custom Metrics

You can add your own custom metrics and have usdmeasureperformance.py include them as part of the set of metrics that it measures.

To define a custom metric, create a script file that defines a testUsdviewInputFunction() function that will be passed to testusdview. For example, if you wanted to add a metric named “process prims”, that traversed the stage and processed each prim in some way, you might have a processPrimsMetric.py script that looks something like:

from pxr import Usd, UsdUtils, Usdviewq

def testUsdviewInputFunction(appController):
    with Usdviewq.Timer("process prims", True):
        stage = appController._dataModel.stage
        for prim in stage.Traverse():
            # process prim as needed, etc

See also the “traverse stage” example in pxr/extras/performance/explicitMetrics/stageTraversalMetric.py.

To include your custom metrics when running usdmeasureperformance.py, add your metrics script name and metric name as part of the --custom-metrics script parameter. For example, if you wanted to include the “process prims” metric example for “MyTestAsset.usda”, you would use a usdmeasureperformance.py command line similar to:

python usdmeasureperformance.py MyTestAsset.usda --custom-metrics processPrimsMetric.py:'process prims'

usdmeasureperformance.py will look for your custom metric script relative to the directory from which the usdmeasureperformance.py script is run.