Benchmarks

The application version number applies to the software as a whole. It changes often as we update the application to add new features and ensure compatibility with the latest hardware. We recommend using the latest application version.

A benchmark version number—for example, Photo Editing Benchmark v1.0—is specific to a test. Benchmark version numbers change rarely and only when absolutely necessary to accommodate changes in third-party applications or bug fixes.

UL Procyon v2.10.1729 — May 2, 2025

This update adds support for the Qualcomm AI Engine Direct (QNN) inference engine and AMD-optimized ONNX models to the AI Image Generation Benchmark. It updates the Intel OpenVINO inference engine to version 2025.1 for the AI Image Generation Benchmark and AI Computer Vision Benchmark. The Qualcomm SNPE inference engine is updated to v2.33 for the AI Computer Vision Benchmark.

This update brings a significant update to the Office Productivity Benchmark PowerPoint score calculation following the deprecation of the PowerPoint Merge function in Microsoft Office version 2025.

AI Image Generation Benchmark v1.1.206.0

New

• Added support for Qualcomm AI Engine Direct (QNN) (Stable Diffusion 1.5, INT8)
• Added support for AMD-Optimized ONNX models (Stable Diffusion 1.5, FP16 and Stable Diffusion XL, FP16)

Updated 

• Updated OpenVINO to 2025.1
• Added support for Qualcomm AI Engine Direct (QNN) (Stable Diffusion 1.5, INT8)

AI Computer Vision Benchmark v1.9.496

Updated 

• Updated Qualcomm SNPE to version 2.33
  • Added support for zero copy memory optimization for the Qualcomm SNPE inference engine. This update results in notable performance improvements, increasing scores by roughly 13% for supported hardware
• Updated Intel OpenVINO to v2025.1

Office Productivity Benchmark v1.5.1397

This update brings a significant update to the Office Productivity Benchmark PowerPoint score calculation following the deprecation of the PowerPoint Merge function in Microsoft Office version 2502. This meant that Office Productivity scores from previous versions of Procyon using Microsoft Office v2502 or later produced lower scores than with previous versions of Microsoft Office.

Please note that this update only changes the constant to make score magnitudes more in line with previous scores. Office Productivity scores from Procyon 2.10.1729 and later are not comparable with scores from previous versions.

Updated

Office Productivity Benchmark

• Updated the Office Productivity Benchmark score calculation following the deprecation of the Microsoft PowerPoint Merge function.
  • This change results in increases for the overall Office Productivity Benchmark Score of ~4%, and an increase for the PowerPoint section score of ~18% compared to Office Productivity Benchmark version 1.4, using Microsoft Office version 16.0.18623.20156, 64-bit.

Office Productivity Benchmark Mult-Platform (MP)

• Updated the multi-platform version of the Office Productivity Benchmark to produce results that better line up with results from the multi-platform Office Productivity Benchmark on macOS. These changes improve scores on windows devices by ~15.5%, with the Microsoft Word section seeing improvements of ~53%. The variation in performance was due to different approaches between operating systems when updating the screen for the following operations in Microsoft Word.
  • Screen updating is now disabled in the benchmark on Windows to match behavior on macOS using AppleScript.
    • add image
    • edit image
    • copy paste
    • cut paste
    • copy from excel
    • table of content
  • Updated the metric measurements for the Microsoft Word section to better match behavior in macOS. These are the metrics that are affected:
    • add image
    • copy paste
    • cut paste
    • table of contents
• Added a script to disable Office Performance Monitoring during benchmarking.
  Some devices can be randomly selected by Microsoft to enable this feature, which may notably affect benchmark results. You can ensure this feature is disabled by performing the following:
  • Opening Task Scheduler
  • Finding the task ‘Microsoft/Office/Office Performance Monitor’.
  • Check if its status is set to ‘Enabled’
  • With Admin privileges, run the script “disable_office_performance_monitor.ps1”.
  • Check the ‘Microsoft/Office/Office Performance Monitor’ task again to confirm it is now set to ‘Disabled’.

UL Procyon v2.10.1663 — Mar 3, 2025

This update adds NVIDIA RTX 50 series GPU support to the TensorRT runtimes in Procyon AI Computer Vision and Procyon AI Image Generation Benchmarks. It also includes updates Intel OpenVINO inference engines for all AI Benchmarks and for the ONNX Runtime in Text Generation Benchmark. 

Along with TensorRT v10.8, CUDA graphs are now enabled in the Procyon AI Computer Vision Benchmarks, bringing notable improvements and scaling to benchmark results on supported hardware. 

We recommend you benchmark your devices again after an inference engine runtime update for up-to-date performance results. 

Please note that after this update, the first AI Text Generation Benchmark run using the OpenVINO runtime will take an extended amount of time, as the benchmark will delete AI models on disk and download models re-quantized with the latest runtime update tools.

Updated

• Updated the Office Productivity Workload to version 1.4.
  • Due to the deprecation of the merge function in Microsoft Office PowerPoint, and the change in PDF export quality, please note that Office Productivity results from previous versions of Procyon on Windows or macOS are not comparable with new results.In our testing sample with Office v16.0.18429.20158 en-US (64 bit), the main Office Productivity Scores were reduced by an average of ~5% for the standard benchmark while Overall PowerPoint Scores were reduced by an average of ~14% . 
• The Intel OpenVINO inference engine has been updated to version 2025.0 for the AI Computer Vision, AI Image Generation, and AI Text Generation Benchmarks  
• The ONNX Runtime inference engine has been updated to version 1.20.1 for the AI Text Generation Benchmark  
• The Microsoft WinML inference engine has been updated to version 1.20.1 for the AI Computer Vision Benchmark  
• The NVIDIA TensorRT inference engine has been updated to version 10.8 for the AI Image Generation and AI Computer Vision Benchmarks. 
  • This TensorRT update enables CUDA graphs for the Computer Vision Benchmark and substantially improves Computer Vision Benchmark scores for supported hardware.
  • This TensorRT update uses quantized models in the QDQ format for the Computer Vision Benchmark because implicit Integer execution via calibration cache is deprecated in TensorRT and unsupported in RTX 50-series.  
• The Procyon Battery Life Benchmark Video Playback scenario option to run the test using the “Movies and TV” application is unavailable for systems running the latest version of Windows 11.  
• Added support for systems running on Arm SoCs in SystemInfo.  

Fixed

• Fixed an issue where the Text Generation Benchmark would not run if the system was offline after models had been downloaded to the disk.  
• CUDA graphs are now re-enabled in the AI Image Generation Benchmark when using NVIDIA TensorRT.  
• A minor discrepancy was fixed in the Computer Vision Benchmark measurements. OpenVINO and Ryzen AI runtimes now include the time taken to fetch the output, the change in performance is negligible.  

Office Productivity Benchmark v1.4

Please note that Office Productivity results from previous versions of Procyon on Windows or macOS are not comparable with new results.

• The Microsoft PowerPoint ‘merge’ step has been removed due to planned feature deprecation in Microsoft PowerPoint for Windows.  This change does not affect the Multi-Platform (MP benchmark). 
• Adjusted the PowerPoint ‘Export to PDF’ quality setting to “slides” instead of “handouts”. This change slightly increases the export time.  This change affects both the standard and  the Multi-Platform (MP benchmark).

Battery Life Benchmark - ‘video playback’ scenario

• The Procyon Battery Life Benchmark Video Playback scenario option to run the test using the “Movies and TV” application is unavailable for systems running the latest version of Windows 11, the scenario is run with the default Media Player application.  
• The option to run the scenario using “Movies and TV” is still available to allow for systems running the older versions of the application. 

UL Procyon v2.10.1542 — December 20, 2024

This update supports the addition of RyzenAI to the Procyon AI Computer Vision Generation Benchmark.

New

• Added support for Ryzen AI in the AI Computer Vision Benchmark. Currently this supports AMD XDNA 2 NPUs.  

Fixed

• Fixed an issue with the AI Text Generation Benchmark runs failing when Procyon was installed on a drive other than “C:”  
• Fixed an issue with the OpenVINO runtime in AI Image Generation Benchmark for NPUs where cached models from an older driver version sometimes affected performance.  

UL Procyon v2.10.1523 — December 9, 2024

This update supports the release of the Procyon AI Text Generation Benchmark. Please note that updates to inference engine runtimes  can impact results. We recommend you benchmark your devices again after an inference engine runtime update for up-to-date performance results.

New

• Added the AI Text Generation Benchmark  

Updated

• Updated TRT integer cache to ones generated with TensorRT 10.6 in Computer Vision.  

Fixed

• Fixed an issue in the AI Image Generation Benchmark where the benchmark would sometimes fail to end correctly when using the OpenVINO runtime.  

UL Procyon v2.9.1481 — November 26, 2024

Please note that updates to inference engines runtimes can impact results. We recommend you benchmark your devices again after an inference engine runtime update for up-to-date performance results.

Updated

• Updated the Intel OpenVINO inference engine to version 2024.5 for the AI Computer Vision and AI Image Generation Benchmarks.
• Updated the Microsoft Windows ML inference engine to version 1.20 for the AI Computer Vision Benchmark.
• Updated the ONNX runtime with DirectML inference engine to version 1.20.0 for the AI Image generation benchmark. DirectML EP now uses the compute command buffer for inference preventing TDR issues on lower-end hardware.
• Updated the Qualcomm SNPE inference engine to version 2.28 for the AI Computer Vision Benchmark.
• Updated the Nvidia TensorRT inference engine to version 10.6 for the AI Computer Vision Benchmark.

UL Procyon v2.9.1429 — November 1, 2024

This update supports the release of the Multi-Platform Office Productivity Benchmark on Windows, letting you compare performance with results from the Office Productivity Benchmark for macOS versions 2.1.250.

Please note that results from the new Multi-Platform Office Productivity Benchmark are not comparable with ‘MP’ scores from previous versions of the Office Productivity Benchmark on Windows. Remember to update the workloads after you perform the UI update to Procyon v2.9

New

• Added the Multi-Platform Office Productivity Benchmark. This can be compared with results from the Office Productivity Benchmark for macOS versions 2.1.250 and later.

Updated

• Added support for Adobe Lightroom Classic 14 in the Procyon Photo Editing Benchmark.
• Added support for Adobe Premiere Pro 25 in the Procyon Video Editing Benchmark.
• Increased the timeout window for the Procyon Photo Editing Benchmark in the ‘Enhance details’ step.

UL Procyon v2.8.1352 — September 16, 2024

This update improves the TensorRT engine build time on first benchmark runs. These updates may impact results for systems using NVIDIA hardware. Remember to check and update the individual benchmarks from the options page or benchmark page after updating the UI. We recommend you benchmark your devices again for up-to-date performance results.

New

• Updated the AI Image Generation workload version to 1.1.144. This adds support for NVIDIA timing caches, improving the speed of building the TensorRT engine on the first run.     

Fixed

• Fixed an issue in the AI Image Generation Benchmark where some systems would fail to build the TensorRT engine or perform poorly on the first run using Stable Diffusion 1.5.
• Fixed an issue in the AI Image Generation benchmark where the benchmark would fail during the model initialization step when using Stable Diffusion 1.5 and the TensorRT inference engine.
• Fixed an issue in the AI Image Generation Benchmark where the benchmark would occasionally fail when run offline.

UL Procyon v2.8.1336 — September 6, 2024

This update supports the release of the integer quantized (INT8) UL Procyon AI Image Generation Benchmark. Remember to check and update the individual benchmarks from the options page or benchmark page after updating the UI. Please note that there are updates to inference engines runtimes that may impact results. We recommend you benchmark your devices again after an inference engine runtime update for up-to-date performance results.

New

• Added support for INT8 quantization for Stable Diffusion 1.5 to the Procyon AI Image Generation Benchmark. This currently supports Intel NPUs, iGPUs and Intel Arc discrete GPUS using OpenVINO, and NVIDIA RTX 30 series GPUs and later using TensorRT.

Updated

• Updated the ONNX runtime with DirectML inference engine to 1.19 for the Procyon AI Image Generation Benchmark  
• Updated the TensorRT inference engine to version 10.3 for the Procyon AI Image Generation Benchmark

Fixed

• Fixed an issue where the Procyon updater was not able to update benchmark workloads.
• Fixed an issue where the Stable Diffusion 1.5 model used in the Procyon AI Image Generation Benchmark was not available. It now uses an alternative Stable Diffusion 1.5 model (nmkd/stable-diffusion-1.5-fp16). This does not affect results Please note that manually installed Stable Diffusion Models need to be removed first, as this new model has a different folder structure.  

UL Procyon v2.8.1305 — August 26, 2024

Please note that updates to inference engines runtimes can impact results. We recommend you benchmark your devices again after an inference engine runtime update for up-to-date performance results.

Updated

• Updated the Intel OpenVINO inference engine to version 2024.3 for the AI Computer Vision and AI Image Generation Benchmarks.
• Added support for OpenVINO NPU Turbo Mode to the AI Computer Vision Benchmark
• Added support for Adobe Lightroom Classic version 13.5 in the Photo Editing Benchmark.

Fixed

• Fixed an issue with Battery Life benchmarks being unable to make a result file at the end of the run. 
• Fixed the VAE step in the AI Image Generation Benchmark to run at FP16 precision when using OpenVINO inference engine, for both Stable Diffusion 1.5 and Stable Diffusion XL. 

UL Procyon v2.8.1251 — July 23, 2024

Please note that updates to inference engine runtimes impact results and should not be compared to results from previous workload versions. Results from the AI Inference Benchmark for Windows from Procyon v2.8.1251 or later are not comparable with results from earlier versions of Procyon.

Updated

• Overall runtime reduced from 3 minutes to 1 minute per model. This does not affect performance.
• Scheduled update for the AI Computer Vision Benchmark TensorRT runtime from 8.6.0 to 10.1.27
• Scheduled update for the AI Computer Vision Benchmark WinML runtime from 1.17.1 to 1.18.1. WinML FP16 and FP32 models have been re-converted using the 1.18.1 ONNX model conversion tool.
• Scheduled update for the AI Computer Vision Benchmark OpenVINO runtime from v2024.0 to v2024.2. OpenVINO FP16 and FP32 models have been re-converted using the 2024.2 model conversion tool.  This patch also explicitly performs a deferred memory copy to make the input visible to the target device and optimizes performance by loading input data directly into device-visible memory. 

Fixed

• Fixed an issue in the Image Generation Benchmark where the UI did not correctly display the selected inference Device for the ONNX runtime. 
• Fixed an issue where Image Generation ONNX runtime models would not be saved locally after being converted.  
• Fixed an issue in the AI Computer Vision Benchmark where the TensorRT inference engine was incorrectly available for ARM64 Qualcomm devices.

UL Procyon v2.8.1207 — June 19, 2024

This is a minor update, benchmarking scores are not affected

Improved

• Optimized startup time for the UL Procyon client.

Fixed

• Fixed an issue where the Video Editing Benchmark would not produce a result at the end of a benchmark run when ‘GPU acceleration’ is enabled.

UL Procyon v2.8.1201 — June 06, 2024

This update supports the release of the One-Hour Battery Consumption Benchmark and the Multi-Platform Battery Life Benchmark. Other benchmark results are not affected

New

• Added the One-Hour Battery Consumption Benchmark
• Added the Multi-Platform Battery Life Benchmark for Windows and macOS  

Fixed

• Fixed an issue in the Computer Vision Benchmark where the Real-ESRGAN test would fail to run when using the WinML inference engine on an intel iGPU.

Updated

• Added support for Adobe Lightroom Classic version 13.3 in the Photo Editing Benchmark.

UL Procyon v2.7.1169 — May 27, 2024

This is a minor update that adds support for Adobe Premiere Pro version 24.2.1 for the UL Procyon Video Editing Benchmark. This update also introduces a toggle for measuring video editing performance with GPU acceleration enabled, producing a separate score. 

In order to support the latest version of Adobe Premiere Pro, benchmark scores are notably affected for and are not comparable with results from UL Procyon Video Editing Benchmark version 1.2.404 or earlier.  

New

• Added the ability to choose between supported AI accelerators for the AI Computer Vision Benchmark when using the Nvidia® TensorRT™ AI inference engine.   
• Added a toggle for enabling GPU acceleration in the Video Editing Benchmark. These scores are not comparable with default Video Editing Benchmark results. Please see the 'UL Procyon Video Editing Benchmark GPU acceleration Toggle' section below for additional context.
• The command line interface now supports using license files for license activations through authenticated proxies.  
  • Please see this article for more information:
    https://support.benchmarks.ul.com/support/solutions/articles/44002125233-activate-a-ul-procyon-benchmark-license-from-the-command-line
• The AI Image Generation Benchmark definition file (for running the benchmark using the command-line) has been separated into multiple files - one for each inference engine.
• The AI Image Generation Benchmark definition files now use device_id instead of device_type, allowing for a more precise selection of available inference accelerators. Please check the definition files for more information on the default values for device selection as well as how to query devices and change them.  
    

Fixed

• Fixed an issue where the device name was missing from AI Image Generation Benchmark results when run from the command-line and using the ONNX runtime AI inference engine.  
  
  

UL Procyon v2.7.1108 — April 08, 2024

This update supports the release of the AI Computer Vision Benchmark for Apple Mac, with the addition of support for the Apple Core ML Inference Engine. Benchmark results are not affected.

New

• Added support for Apple Core ML Inference engine results.

Updated

• Updated the AI Computer Vision workload to version 1.5.290
• The ResNet50 AI model has been replaced with a PyTorch-based version, as the previous ONNX based ResNet50 version could not be converted into Core ML. This does not meaningfully affect benchmark results.

UL Procyon v2.7.1106 — April 04, 2024

This is a minor update, benchmark results are not affected

New

• Added a TensorRT device selector to the AI Image Generation Benchmark for when the system has multiple supported NVIDIA GPUs.
• Added new warning labels to the AI Image Generation Benchmark for systems that fail to meet the minimum system requirements.

Fixed

• Fixed an issue in the Image Generation Benchmark where the benchmark would fail when run in offline mode, even though the AI models were downloaded onto system storage.
• Fixed an issue where the benchmark would not correctly use FP32 precision with OpenVINO on some hardware configurations.
• Fixed an issue in the AI Computer Vision Benchmark where the WinML inference engine was incorrectly available for ARM64 Qualcomm devices.
• Fixed an issue in the AI Image Generation Benchmark where failed benchmark runs would incorrectly produce an overall score instead of a 0.
• Fixed an issue where systems using Intel NPU hardware would be able to select the FP32 precision setting for the OpenVINO inference engine. Current Intel NPU hardware does not support this precision inference. 

Updated

• Updated the AI Computer Vision Benchmark workload to version 1.4.289
• Updated Qualcomm SNPE to version 2.20
• Updated Intel OpenVINO to version 2024.0
• Updated Microsoft Windows ML to version 1.17.1
• Updated SystemInfo to version 5.71

UL Procyon v2.7.1096 — March 25, 2024

This update supports the release of the UL Procyon AI Image Generation Benchmark, other benchmark results are not affected

New

• Added the UL Procyon AI Image Generation Benchmark

Fixed

• Fixed an issue where the UL Procyon application could not parse system info on some system configurations.

Updated

• With the new AI Image Generation Benchmark being added to the UL Procyon AI Inference Benchmarks family, we’re clarifying our existing AI Inference Benchmark for Windows as the UL Procyon AI Computer Vision Benchmark.

UL Procyon v2.6.1059 — March 06, 2024

This is a minor update, benchmark results are not affected

Fixed

• Fixed an issue where results exported to PDF would be missing hardware monitoring graphs.
• Fixed a rare issue where a “Visual C++ Runtime install failed with error code 1638” error would display, despite the update completing successfully.

New

• Added support for UL Procyon to connect to the internet through proxy to confirm licenses, update, and submit benchmark results.

Updated

• Files and documents are now prelabelled with “Public” sensitivity labels for environments that require documents to have a mandatory label.
• Updated SystemInfo to version 5.70.  
  Starting with this version, CPU load monitoring data is now based on the CPU utilization metric, making it consistent with the Windows Task Manager. Please note that this means CPU load monitoring graphs from versions 5.70 and onwards are not visually comparable with graphs from results using earlier versions of SystemInfo.
  Benchmark results are not affected by this change.
  Please note: the CPU utilization metric can now exceed 100%. For more information, please see this Microsoft troubleshooting article.

UL Procyon v2.6.988 — January 17, 2024

This is a minor update, benchmark results are not affected

Updated

• Updated the Procyon AI Inference Benchmark to show the correct version number of AI Workload version 1.3.276 
• Fixed an issue where the ‘-export-pdf' command line option would fail with a ‘failed to write pdf’ error.

UL Procyon v2.6.963 — January 04, 2024

This version updates the Intel OpenVINO, Qualcomm SNPE and Microsoft WindowsML runtimes. Updates to SNPE and OpenVINO result in performance improvements on supported hardware.

Results from the AI Inference Benchmark for Windows from Procyon v2.6.963 or later are not comparable with results from earlier versions of Procyon.

New

• Added support for Qualcomm Snapdragon X Elite SoCs.

Fixed

• Fixed an issue where the UI toggle for SystemInfo hardware monitoring would become unresponsive.

Updated

• Updated the AI Inference Benchmark for Windows to version 1.3.276

Please note: This patch does not display the correct workload version for the AI Inference Benchmark for Windows in the UI. The workload shows version 1.2.273 instead of the correct 1.3.276. 

AI Inference Benchmark for Windows runtime update - Version 1.3.276

• Updated Intel OpenVINO inference engine to v2023.2. This runtime update increases scores by up to 5% for most supported hardware, with some components seeing more significant score improvements.
• Updated Qualcomm SNPE inference engine to v2.17. This runtime update increases scores by roughly 10% on supported hardware.
• Updated Microsoft WindowsML inference engine to v1.16.3. This version requires DirectML v1.13.0 or later.

UL Procyon v2.6.931 — November 13, 2023

This is a minor update. This version updates the Intel OpenVINO runtime, which results in performance improvements on supported hardware.

New

• Added support for testing the inference performance of NPUs supported by the Intel OpenVINO AI Inference Engine version 2023.1.

Updated

• Updated Intel OpenVINO AI Inference Engine to 2023.1. This runtime update increases scores by roughly 5% for most supported hardware, with some components seeing more significant score improvements.
• Qualcomm SNPE workload changed to use the C API instead of C++ API, as Qualcomm is retiring the C++ API. Benchmark scores are not affected.

UL Procyon v2.6.920 — October 30, 2023

This is a minor update. Benchmark scores are not affected.

New

• The simple-csv results export option now also exports the following fields: Procyon version, OS version, Application version, AI inference engine version.

Updated

• Transitioned Java Runtime from Oracle v1.8.0_91 to OpenJDK v11.0.20.
• Added Procyon Photo Editing and Video Editing Benchmarks support for Adobe Photoshop 25.0, Lightroom Classic 13.0, and Premiere Pro 24.0.

Fixed

• Fixed an issue where the Procyon AI Inference Benchmark for Windows would sometimes incorrectly run on the CPU when a dedicated GPU was selected when using the OpenVINO AI inference engine.
• Fixed an issue where the language selection dropdown would not show any language options.

UL Procyon v2.6.848 — August 03, 2023

This is a minor update. Benchmark scores are not affected.

Fixed

• Fixed a rare issue in the AI Inference Benchmark for Windows where tests using the Windows ML runtime would use the Microsoft Basic Render driver instead of the selected GPU.

UL Procyon v2.6.847 — July 27, 2023

This update includes updates to the runtimes used in the Procyon AI Inference benchmark for Windows. Updates to AI runtimes changes performance. Results from the AI Inference Benchmark for Windows from Procyon v2.6.847 or later are not comparable with results from earlier versions of Procyon.

AI Inference Benchmark for Windows runtimes

• TensorRT updated to version 8.6.1 GA 
• Windows ML updated to version 1.15.1 (DirectML to version 1.12.0)
• OpenVINO updated to version 2023.0
• OpenVINO models updated to IR11

New

• New FP16 precision option for Windows ML, OpenVINO and TensorRT runtimes 
• New GPU selector for Windows ML runtime 
• New device selector for OpenVINO runtime

UL Procyon v2.5.829 — June 21, 2023

Fixed

• Fixed an issue where AI Inference Benchmark for Windows would not use the selected component for running the test.
• Fixed an issue where the Battery Life Benchmark video playback scenario would not correctly report version information for the Microsoft Movies & TV / Microsoft Films & TV App.

Improved

• When running Procyon via the command line with the --loop command, exporting results to a CSV file will now export results from all loops into a single CSV file, instead of results from only the first loop.

UL Procyon v2.5.798 — May 4, 2023

This update supports the launch of the office productivity scenario for the Procyon Battery Life benchmark. Other benchmark scores are not affected. 

New

• Added the office productivity scenario for the Battery Life benchmark.

Fixed

• Fixed an issue where the Photo Editing benchmark would fail at the enhance details step in Lightroom Classic 12.3.

UL Procyon v2.4.783 — April 27, 2023

This is a minor update. Benchmark scores are not affected. 

New

• Added support for MP Scores in the office productivity benchmark.

UL Procyon v2.3.745 — April 20, 2023

This is a minor update. Benchmark scores are not affected. 

Fixed

• Fixed an issue that prevented exporting results to XML from the results screen.

UL Procyon v2.3.713 — March 22, 2023

This update supports the launch of the new UL Procyon AI Inference Benchmark for Windows. Other benchmark scores are not affected.  

New

• Added the UL Procyon AI Inference Benchmark for Windows. 
• Added the screen calibration tool to the UL Procyon Battery Life Benchmark. 
• Added additional system information to the UL Procyon Battery Life Benchmark results screen.

Fixed

• Fixed an issue where the UL Procyon Office Productivity Benchmark would not complete due to a “no permission” fail in the Outlook step. 

UL Procyon v2.2.699 — March 07, 2023

This is a minor update. This update includes changes to benchmark elevation state that may affect scores on systems using 9th generation Intel CPUs or older.

Elevation requirement changes

• Removed elevation request when starting the UI.
• Benchmarks will always run at a specific elevation, independent of UI elevation state.
  • UL Procyon Video Editing Benchmark must be run as elevated.
  • UL Procyon Office Productivity, Photo Editing, and Battery Life Benchmarks will always run not elevated.   

Improvements

• UL Procyon Suite UI updated to version 2.2.349. It now targets .NET 4.8 and Visual Studio C++ 2022.
• Updated all tests to target .NET 4.8 and Visual Studio C++ 2022.

Tests

• UL Procyon Office Productivity Benchmark updated to version 1.1.
• UL Procyon Photo Editing Benchmark updated to version 1.1.
• UL Procyon Video Editing Benchmark updated to version 1.1.
• UL Procyon Battery Life Benchmark updated to version 1.1.

UL Procyon v2.1.657 — January 23, 2023

This is a minor update. Benchmark scores are not affected. 

New

• Added the option to import multiple UL Procyon licenses from a single file.

UL Procyon v2.1.544 — October 27, 2022

This is a minor update. Benchmark scores are not affected. 

Compatibility

• The Video Editing Benchmark is now compatible with Adobe Premiere Pro v23.
• The Photo Editing Benchmark is now compatible with Adobe Lightroom Classic v12.

UL Procyon v2.1.527 — October 10, 2022

This is a minor update. Benchmark scores are not affected. 

Fixed

• Fixed an issue where the Office Productivity benchmark would fail on some systems during the PowerPoint test PDF export step.

UL Procyon v2.1.459 — July 6, 2022

This update fixes an issue in the face detection part of the Photo Editing Benchmark that sometimes resulted in incorrect scores. 

Fixed

• Fixed an issue in the face detection part of the Photo Editing Benchmark where, in some cases, face detection did not start, causing the final benchmark score to be incorrect. This issue affects UL Procyon versions 2.0.325, 2.0.360, 2.0.399 and 2.1.457. Photo Editing Benchmark results from these versions with a "Face detect" result time less than 5 seconds are invalid and should not be used for comparisons.
• The Photo Editing Benchmark now supports Polish and Norwegian language settings in Adobe Lightroom Classic.

UL Procyon v2.1.457 — June 30, 2022

This major update adds the UL Procyon Battery Life Benchmark for measuring the battery life of Windows laptops, notebooks and tablets.

New

• UL Procyon Battery Life Benchmark.
• You can now submit, save and manage your Procyon benchmark results on 3dmark.com. You can choose whether your results are public or private.

Improved

• Improved layout and design of benchmark details screen.

UL Procyon v2.0.449 — June 21, 2022

This is a minor update. Benchmark scores are not affected. 

Fixed

• Results from Intel 12th generation “Alder Lake” processors with E-cores no longer have missing hardware information when exported to PDF.

UL Procyon v2.0.399 — April 4, 2022

This is a minor update. Benchmark scores are not affected. 

New

• The UL Procyon app is now available in Japanese. You can change the language setting on the Options screen.

Fixed

• The benchmark result PDF now correctly identifies systems with Windows 11.

UL Procyon v2.0.355 — February 21, 2022

This is a minor update. Benchmark scores are not affected. 

Fixed

• The benchmark result screen now correctly identifies systems with Windows 11.

UL Procyon v2.0.325 — February 4, 2022 

This is a minor update. Benchmark scores are not affected.  

Improved 

• Updated the face detection part of the Photo Editing Benchmark to improve cross-platform score comparisons with the upcoming MacOS version of the benchmark. 

Fixed 

• The Photo Editing Benchmark now produces an error notification if the enhance details part of the test fails to run. 

Compatibility 

• The Video Editing Benchmark is now compatible with Adobe Premiere Pro 22.3.0 beta build 29. 

UL Procyon v2.0.324 — January 21, 2022

This is a minor update. Benchmark scores are not affected. 

Improved 

• You can now open a result with a single click on the Results screen.  
• Added the benchmark version number to the benchmark details screen. 
• Various interface and layout improvements. 

UL Procyon v2.0.303 — December 21, 2021

This is a minor update. Benchmark scores may be affected. The Outlook “new appointment” metric may be up to 5% faster. As a result, the Outlook test score may be up to 1% higher, and the overall Office Productivity Benchmark score may be up to 0.5% higher.

Fixed

• Modified the Outlook test for improved compatibility with enterprise environments with system policy settings and user access permissions that previously caused the Outlook test to fail when trying to save an appointment.

UL Procyon v2.0.289 — November 25, 2021

This is a minor update. Benchmark scores are not affected.

Fixed

• Fixed Outlook version number detection.
• Fixed an issue resulting from Microsoft Word’s AutoRecover feature that could cause the Word test to fail to run.

UL Procyon v2.0.279 — November 16, 2021

New  

• The UL Procyon app is now available in Brazilian Portuguese, German, Simplified Chinese and Spanish. You can change the language setting on the Options screen.

UL Procyon v2.0.276 — November 3, 2021

This is a minor update.

UL Procyon benchmarks use real applications whenever possible. Updates to those applications can affect your benchmark score.

Compatibility

• The UL Procyon Video Editing Benchmark is now compatible with Adobe®️ Premiere️®️ Pro 22. 
• The UL Procyon Photo Editing Benchmark is now compatible with Adobe®️ Lightroom®️ Classic 11.0. Please note that exporting an image takes longer in Lightroom Classic 11.0 than in 10.x. Photo Editing Benchmark scores from systems with Lightroom️ Classic 11.0 will be lower than scores from systems with Lightroom️ Classic 10.x.

UL Procyon v2.0.268 — October 29, 2021

This is a minor update. Benchmark scores are not affected. 

New   

• Added Virtualization-based Security (VBS) status to the System Information section of the benchmark result screen. 

Improved  

• Added the UL Procyon AI Inference Benchmark for Android devices to the My Suite screen. 
• Exported PDF results from the Office Productivity Benchmark now include the workload sub-scores. 
• The Options screen now displays the UL Procyon application version number. 

Fixed 

• Fixed an issue that could result in incomplete monitoring data for some benchmark runs. 
• Added a workaround to prevent pop-up windows from Microsoft Word’s AutoRecover feature interrupting the Office Productivity Benchmark.  
• The Options screen now shows the Video Editing Benchmark version number when the application is in an unlicensed state. 
• Fixed the “Write detailed log”’ setting on the Options screen. 

Known issues 

• The UL Procyon Photo Editing Benchmark is not yet compatible with Adobe®️ Lightroom®️ Classic 11.0. This will be fixed in the next update. Please use Adobe️ Lightroom️ Classic 10.x if possible. 
• The UL Procyon Video Editing Benchmark is not yet compatible with Adobe®️ Premiere️®️ Pro 22. This will be fixed in the next update. Please use Adobe️ Premiere️ 15.x or 14.5 or later if possible.

UL Procyon v2.0.249 — October 18, 2021 

This is a minor update. Benchmark scores are not affected. 

Fixed 

• Fixed an issue that could cause the Photo Editing Benchmark to return a score of zero as the result. 

UL Procyon v2.0.248 — October 11, 2021

This is a minor update. Benchmark scores are not affected.

Improved  

• Added links to the UL Procyon user guide to several screens in the app. 
• Added a support section to the Options screen. 
• Changed the “Delete” option on the Results screen to “Remove” to make clear that results are only removed from the table view and not deleted from the system. 
• You can now expand and collapse storage drive details in the System Information section of the result screen. 

Fixed 

• The temporary backup file used in the Outlook test is now removed after running the benchmark when Outlook is using a language other than English. 
• Added a missing status message to the benchmark progress panel.  
• Added missing labels for storage device details to the System Information section of the result screen. 
• Removed the nonfunctioning “Export as Excel” option from the Results screen. This feature is not yet ready for release. 
• Removed the “Highlight differences” toggle from the Result Comparison screen. This feature is not yet ready for release. 

Known issues

• The Photo Editing Benchmark may return a score of zero in some cases. This will be fixed in the next update.

UL Procyon v2.0.232— October 4, 2021 

This major update adds the Office Productivity Benchmark and a redesigned user interface for the Procyon application. 

New  

• UL Procyon Office Productivity Benchmark. 
• Redesigned application user interface. 
• You can now compare up to four results side by side in the app.

UL Procyon v1.0.183 — March 17, 2021

Compatibility

• UL Procyon Photo Editing and Video Editing Benchmarks can now use Adobe Creative Cloud applications in all supported languages. The previous version required English-language Adobe apps.
• The UL Procyon Video Editing Benchmark is now compatible with Adobe Premiere Pro v.15.0.

UL Procyon v1.0.180 — January 29, 2021

Fixed

• Fixed an issue that prevented the Photo Editing Benchmark from opening Adobe Lightroom Classic v10.1.1.

UL Procyon v1.0.172 — November 30, 2020

• Launch version with Photo Editing Benchmark and Video Editing Benchmark.