PassMark PerformanceTest - What's New?
Here’s a summary of all the changes that have been made in each version of PerformanceTest. Click here to return to the PerformanceTest product page.
V9.0 Build 1000 - 26th/September/2016
- New DirectX 12 test that runs at a default resolution of 4K (3840x2160). In cases where 4K resolution isn't available the score will be scaled down and the test will run at the highest available resolution.
- Removed simple DirectX 9 test from 3D test suite. This was removed as DX9 is now fairly old and the simple test wasn't putting much load on newer video cards.
- CPU extended instructions test will perform testing using sub-tests for FMA, AVX and SSE (or only those that are supported) and take the average of the 3 (or of those that are supported) for the benchmark result. FMA and AVX are new CPU instruction sets that have become available over the last few years in Intel and AMD CPUs. They were designed to make certain mathematical operations faster. Individual sub-test duration has also be adjusted to account the new sub-tests.
- CPU Physics Test and the Advanced Physics Test have been switched to use the current 'Bullet' physics engine. In the past we were using the Tokamak engine, but Tokamak is no longer maintained. While at the moment GPU acceleration isn't stable in Bullet, we hope to enable GPU acceleration in Bullet in a future patch (limited to integrated GPUs).
- 2D Graphics, added a Fonts and Text test using Direct2D. This test will launch after the original Font and Text test complete on systems that support Direct2D (Vista and higher). Direct 2D is used by most browser and some other applications for rendering text and web pages. So the new Fonts and Text test result will now be a combination of Direct2D performance and GDI performance. GDI being the traditional method if displaying text and graphics in Windows. While Microsoft advertised Direct2D as being faster than GDI, we found the opposite in most cases. Win10 did bring some Direct2D performance improvements however.
- DirectCompute (GPU Compute) tests will now run for longer based on the test duration configuration setting. The bitonic test will now run for a length of time corresponding to the duration setting, previously it was set to 100 iterations.
- Added an OpenCL test to the GPU Compute set of tests
- GPU Compute bitonic, standard test will launch two instances of bitonic test is there are 2 identical cards in the system (not running in SLI/XFire mode)
- Advanced Disk Test
- Added sampling of individual latency results and added support for latency results plotted on the graph as a heat map.
- New Drive IOPS Test available from "Advanced->Disk->Drive Performance..." menu option.
- Advanced Memory Test
- NUMA is now supported where available. Settings are displayed when selecting results to graph. Exported results and graph will now include NUMA settings.
- Added DX11 Oculus Rift VR test, as an advanced test. This is currently a bit experimental and it has only had limited testing. The test is based on the existing DX11 test (the Lunar landscape test), but has been redesigned slightly to run on an Oculus Rift.
- Advanced 3D Test
- Changed how results were displayed and will now display frame latency times. Note: Frame latency is measured from the applications point of view, so doesn't tell the whole story, as the video card can buffer or discard frames.
- Added button for easy copy of results to clipboard
- Advanced Direct Compute (GPU Compute)
- Renamed to GPU Compute
- Added an OpenCL test
- Added option for bitonic test to run on all available cards. Previously the test would only run on a single card.
- Advanced network test, added internet and transfer protocols used to test report output. Also changed the maximum block size of 256KB and set the default to 32KB. The larger block sizes result in less overhead and higher throughput.
- Advanced Visualized Physics Test
- Now uses Bullet physics engine (see above)
- Added a new collection of preset tests, based on the capabilities of the Bullet engine. Ragdoll, compressible objects, friction, rope simulation, etc..
- Switched to using OpenGL3 (instead of DirectX) for rendering the scene.
- Changed max number of shapes to be 30000 (limited by Bullet Physics)
- Removed Overall FPS display (same as render FPS due to multithreading)
- Made some changes so GPU Compute tests are now DPI aware and should no longer automatically scale on high DPI monitors
User interface changes
- Totally new look user interface, redesigned with new charts for displaying benchmark data, your result vs the world (using all submitted baselines) and your result vs same hardware model (CPU/Video/RAM/Disk).
- Added User editable string file allowing for customisation and translation of strings used in PerformanceTest (pt_localization.txt). Nearly all menus and headings can be customised/translated by editing the text file. This will allow future localization of the software into French, Chinese, etc..
- Added option to enable/disable heat map plotting in Customization dialog (in the advanced tests)
- Added real-time temperature collection and display to main 3D window display
- In system information window physical disks are now displayed in disk number order (as opposed to the SMBIOS order)
- Added baseline drag and drop loading to main baseline information window, baselines can be dragged form windows explorer into PerformanceTest for loading.
- Added right-click menu to benchmark graphs for exporting results to clipboard/file
- Removed "Show numeric results column" from Preferences
- Temperatures are now updated after corresponding test suite has completed (eg. CPU temperatures updated after running CPU test suite, etc)
- When RAM manufacturer/part number from the SPD data is not available, use SMBIOS details instead
- Added mouse over tooltips for advanced memory graphs
- Added DPI scaling for 2D font test. Previously the test window would be tiny on high DPI displays.
- Replaced 'Upload baseline' button with checkbox that is checked by default. Checkbox is enabled only if full baseline is available and baseline has not already been uploaded.
- Replaced some icons with higher quality ones
- When running a single test a reduced size status window is now displayed
- Removed Vsync options in main preferences, will now always default to off. Script command will remain and allow it to be changed when running a script.
- Added German, French, Chinese (simplified and traditional) translation
- Scripting, added REPORTSUMMARYHTML script keyword to output report in HTML format when using the LOOPS command
- Scripting, added optional parameter SINGLELINE to commands EXPORTCSV, EXPORTTEXT for outputting the computer's system information and results to one line
- Added shortened operations descriptions to CSH/HTML/TXT exports (eg M/ops, Mb/s, Fps, etc)
- Added error message when the number of selected baselines cannot fit on one page when printing
- Added error message when saving results to image fails due to the image file height being too large (eg too many selected baselines)
- Added a warning message when submitting a baseline if different video card names were detected as being tested. This is to alert the user and try to prevent situations in systems with low power/high power video cards using the low power card in some tests
- Support for new CPUs and better identification of AMD graphics card (some AMD card have never identified themselves correctly and only refer to themselves as 'Series 200' or the like).
- Added export function to export results and graphs to HTML file from drive performance
- Changed an error message when then DirectX12 test could not be launched
- Added a delay before launching 3D tests when running a script and using their single test command to prevent monitor resolution switching occurring too quickly and causing errors
- Scripts will now exit and no longer continue to run if an error occurs during the SETDISK command
- Changed the behaviour of the Dx9 and DX10 when run from a script, previously they would run for much longer than the duration used for SETDURATION
- Added network drives to Advanced Disk -> Drive Performance disk selection
- Changed scripted CSV output to always print test names and results, if no result recorded then output N/A, to allow for easier combination of results from multiple machines
- Added option to 'Clear chart cache' in Preferences window. All .xml files under ProgramData\Passmark\PerformanceTest9\ChartData\ are deleted.
- Added a check for Nvidia G-Sync, if detected (and if result seems to be capped at a framerate between 54 and 60) will block basline submission until G-Sync is turned off and 3D tests re-run
- Changed how script files are processed to better handle the loop command
- Fixed some places where files were saved to the local user directories instead of the executable directory when running with command line parameter /p (eg not saving to USB drive as expected)
- Fixed a bug in the memory read cached test where the buffer size used was in steps from 16KB to 148KB instead of 16KB to 8MB. Test results in this test will now be slightly lower than they were as a result.
- Fixed baseline search in baseline manager where searches of turbo speed 5001MHz+ were not working correctly
- Fixed a bug in the Advanced Direct Compute bitonic test where the displayed results were being calculated incorrectly
- Added some checks to try and avoid a situation where a graphics driver crash could cause an abnormally high Direct2D result
- Fixed a crash when running Direct Compute test
- Now closing DirectIO.sys library when PT exists to stop the .sys file being in use until automatically closed. This should avoid an error some people saw on installation that the DirectIO.sys was locked and couldn't be overwritten.
- Fixed incorrect values for average speed + data transferred when exporting results from the advanced network test
- Fixed crash in advanced network test if test was started without server being available.
- Fixed a bug where the ADT_SETEXPORT graph settings were not being read correctly and the graph type was not changing
- Fixed a bug where DX11 test would often pick a low refresh rate (eg 24hz)
- Fixed a bug where the "raw disk read" options for the advanced CD/DVD test was unavailable in windows versions newer than XP
- Fixed some CSV output so "," characters in CPU names no longer break the formatting
Test Score Scaling
- Added 6.5% scale down of CPU Mark so PT9 results are closer to PT8 results, to compensate for the increased scores in the Extended Instructions and Physics test changes.
- So the PT8 CPUMark should be broadly comparable to the PT9 CPUMark. Same for the overall PassMark rating. Still, when possible, for improved accuracy, compare PT9 results to PT9 results and PT8 results to PT8 results.
- DX12 test score is penalised if it cannot be run at the desired 4K (3840x2160) resolution. Running at lower resolution increases the frame rate. So the scaling compensates for this.
- The weighting of the 3D tests was changed due to the removal of the DX9 simple test and the addition of the DX12 test. The DX9 (complex) test is now weighted lower than the other tests (approx. 25% less). Systems that cannot run the DX12 test will received a 20% penalty to the overall 3D mark score.
V8.0 Build 1054 - 17th/June/2016
- Fixed a bug that was causing advanced baseline searches to time out
- Added some new checks for disk caching software
V8.0 Build 1053 - 9th/December/2015
- Work around was made to avoid abnormally high Direct2D benchmark results if the graphics card driver crashed during the test. Some versions of Intel's video driver would crash with message "Display driver stopped responding and has recovered". Switching to the Mircosoft supplied video driver also fixed the problem.
- Fixed a bug in the Advanced DirectCompute bitonic test where the displayed results could be calculated incorrectly
- Updated system information library
- Added support for Intel Xeon Broadwell-E CPU
- Fixed Intel Burst mode frequency measurement (it should be noted that Burst mode it not the same as Turbo mode)
V8.0 Build 1052 - 22nd/October/2015
- Updated system information library to recognise SLI and Xfire
- Added SLI field to baselines and info display
- Fixed a crash that could occur on some Intel GMA 3150 / 945 express graphics systems
V8.0 Build 1051 - 6th/August/2015
- Added support for Windows 10 and option to select Windows 10 when searching for baselines
- Updated system information library
- Fixed an issue where graphics cards with large amounts of memory were being displayed as having 4GB only
- Support added for Intel Skylake CPUs
V8.0 Build 1050 - 2nd/July/2015
- System Information updates - Added support for AMD Radeon R9 Fury X, R9 390X, R9 390, R9 380, R9 M375, R9 M370X, R7 370
- Fixed a error message displayed when using the auto launch command line parameter /a and not specifying an initialisation file
V8.0 Build 1049 - 17th/June/2015
- When submitting a baseline removed the need for a 3D mark to be included before submission to make it easier for server style setups with low power video cards to submit baselines
V8.0 Build 1048 - 15th/June/2015
- Fixed a bug on some systems with multiple CPUs where the correct number of CPU test threads was not starting
- Fixed a crash that could happen in the DX11 test
V8.0 Build 1047 - 28th/April/2015
- Changed advanced baseline searches so they should be faster when switching between pages of results
- Added a default "no sorting" option for advanced baseline search to speed up searches
- Fixed a bug with the cursor not always redrawing correctly immediately after a baseline search finishes
- Changed simple search to report "X baselines found" when < 100 or "Showing first 100 baselines" when more are found as we are no longer counting them to increase the speed of the search
- Added a check and removal of BOM when loading a customised key.dat file (file expected to be in unicode UTF-16LE format)
- Fixed broken parsing of pt.ini file when using /a command line options, pt.ini will be read in ascii or unicode (UTF-16LE) format
- System Information updates
- Added support for AMD HD 8690A
- Updated CPU specification information,aAdded Intel Cherry Trail, Braswell and Knights Landing, newer Intel Haswell and Broadwell CPU's, newer AMD CPUs.
- Added Intel Bay trail temperature monitoring.
V8.0 Build 1046 - 11th/March/2015
- Baseline Manager basic Search will now require the user specify a field (CPU / Video / Drive / Motherboard / Description) to search
- Baseline Manager, an Empty Advanced Search (with the Default Options) will result in showing the last day of submissions
- Fixed a crash that could happen when reading in the windows experience performance values in win 8
V8.0 Build 1045 - 27th/January/2015
- Fixed a bug where the "Save baseline" menu item was opening the "Export results as image" dialog
V8.0 Build 1044 - 23rd/January/2015
- Added option to display temperature as Celsius or Fahrenheit
- Made some changes so on some systems with dual Intel/AMD video cards the correct card is displayed in the results
- System Information updates
- Added support for AMD Radeon R9 285, AMD Radeon HD 7730
- Fixed a bug in the "Install to USB" function that would stop the install process when the USB drive ran out of space but didn't display an error message
- Fixed some incorrect tool tips in the File menu
V8.0 Build 1043 - 18th/December/2014
- Changed the memory information displayed on the memory tab when a test has just been run for "This computer" to match the summary and system information displays
- Fixed a bug with the script command REPORTSUMMARY where if no file path was entered the file was not being created in the default user directory
- Fixed a display bug where on systems that had no CD drive connected an overly large result would be displayed when trying to run the CD test
- System Information updates
- Added support for AMD R9 280
- Fixed possible crash due to size of string buffer not being big enough for certain DDR3/DDR4 attributes
- Fixed a crash if a localisation string failed to load
V8.0 Build 1042 - 28th/October/2014
- Added motherboard model name (where available) to PassMark rating information on summary tab
- Made a change so scripts will now process environment variables used for the EXECUTE command
- Replaced reference baseline #5 with a different system due to higher than normal 2D and memory results
- Replaced reference baseline #6 with a higher end system to match current hardware
- System Information updates
- Improved Bay trail supported (Corrected the number of L1 and L2 CPU caches on Bay Trail CPU's, CPU frequency display improvements)
- Intel Silvermont CPU support added
- Added custom names for AMD A10/A8/A6 integrated cards (eg AMD Radeon R7 A10-7850k)
- Added support for AMD Radeon HD 8xxxD/7xxxD/6xxxD/8400/8330/6320 video cards
- Fixed a bug in the advanced disk test where the data randomness value was being flipped each edit (eg 100 - 0)
- Fixed a possible crash that could occur in the DX10 test if some required resources became invalid
V8.0 Build 1041 - 25th/September/2014
- The video card / RAM / disk / CPU tested should now be displayed next to the "This computer" and local baseline names on the results tabs
- Fixed a bug where on some switchable graphics systems (eg Optimus) the video card used was not always displayed next to the baseline number when viewing the 2D and 3D results
- Changed behaviour when PerformanceTest is gathering video card information so it should now display some NVIDIA video cards that were previously hidden. This applies to SLI systems where only one monitor was connected and to Optimus systems.
V8.0 Build 1040 - 18th/September/2014
- Added a PNP Device ID display for graphics cards in the 3D information view
- Changed behaviour on systems with multiple video cards (switchable graphics/SLI/XFire), previously PT might not always display the video card in the system information if they did not have a monitor attached
- Fixed a printing issue caused by the previous high DPI changes
V8.0 Build 1039 - 10th/September/2014
- Some system information updates which enable better DDR4 and Haswell-E support
V8.0.Build 1038 - 29th/August/2014
- Updated system information collection to support Intel broadwell and next generation of Intel Xeon Processor Family based on Haswell
- Enabled "DPI aware" option for manifest file so PT will work correctly in windows 8 high DPI (QHD+) resolutions
- Updated the DX9 3D tests to correctly display information text on high DPI displays
- Replaced several instances of "Mhz" with "MHz" in baseline display text
- Fixed a rounding display error for RAM values in the baseline manager
- Fixed a bug where it was possible to start 2 advanced 3D test windows at the same time which caused PT to crash
- Fixed a bug where scaling of fonts was not happening correctly on QHD+ high DPI displays
- Fixed a bug where incorrect resolution could be displayed on QHD+ high DPI displays in system information
- Fixed a crash that could occur when collecting memory SPD information at startup
V8.0.Build 1037 - 25th/July/2014
- Added check for AMD powerxpress to try and correctly detect AMD graphics cards in dual card laptops
- Added support for AMD R5 M230, R5 M255, R7 M260, R7 M265, R9 M275X, R9 M265X, R9 M270X, R9 M290
- Added DDR4 SPD decoding
- Increased timeout when using the simple search option in baseline management to prevent an issue where no results were being returned
V8.0.Build 1036 - 20th/June/2014
- Fixed a library error that was preventing PerformanceTest from running in Windows XP
V8.0.Build 1035 - 17th/June/2014
- Added a warning on startup when using a "Power saver" power plan
- Added Atom NXXXX Temperature monitoring
- Added AMD R9 295X2 and R7 260 support
- Fixed an issue where in some cases systems were not displaying all their installed RAM modules correctly (more noticeable in systems with >16 RAM slots)
V8.0.Build 1034 - 7th/May/2014
- Added a missing localisation text file to the install package, this was causing some system information to be displayed without formatting
- Made a change to fix an issue on some systems where more RAM modules were displayed than was installed in the 3D main window graphic
- Made some changes to how the DX10 test starts to try and avoid some situations where the test will display a blank window and not start the fullscreen test (in multiple monitor situations mainly) until there was some user input (eg a moving the mouse)
- Made some changes to detect and fix invalid baselines when downloading them
- Fixed a bug with baseline date filtering where the "To" date was not being set properly and was using the current date
- Fixed a bug with baseline date filtering where the "To" date was being set to the "From" date when re-opening the "Date & sorting" dialog
V8.0.Build 1033 - 24th/April/2014
- Fixed a crash (bluescreen) that could occur in the 32bit version when reading system information
- Partitions on hard disks with 4KB sectors can now be tested in the advanced disk test, previously only the raw device was displayed.
V8.0.Build 1032 - 3rd/April/2014
- Changed behaviour of the DX10 test so it should pick the primary adapter and monitor when mutiple monitors are connected to multiple video cards
- Fixed a crash that could occcur on startup when running in a virtual machine.
V8.0.Build 1031 - 11th/March/2014
- Added 4MB and 8MB block sizes to the advanced disk test
- Added SETMEMTIMEOUT script command to allow the threaded memory test timeout length to be changed
- Fixed a bug that could occur when reading memory information which could cause a crash
- Fixed a bug where if an invalid motherboard serial number was written to the baseline file the saved file could not be re-opened
V8.0.Build 1030 - 18th/February/2014
- PerformanceTest should now distinguish between Windows 8 and 8.1 in the system information
- Baseline manager will now display baseline scores with a decimal point. As this value is also used for sorting the displayed list it also fixes a bug where baselines separated by a decimal value (eg 100.0 and 100.1) were being displayed in the wrong order.
- Baseline manager will now check baseline creation time with the time stored in the baseline cache file (baseline.cfg). If the file in disk is more than 30 mins newer than the cache entry the cache is updated. Previously overwriting a file in "My Baselines" would not update the score listed in the baseline manager.
- Fixed an error that would be displayed if a baseline had an invalid temperature value stored in it.
V8.0.Build 1029 - 14th/January/2014
- Disabled baseline uploading when Samsung RAPID service is detected and the disk test has been run on a Samsung hard drive (the tests can still be run but the results cannot be uploaded)
- Disabled mouse and keyboard input for the duration of the 2D tests to precent windows being minimised
- Fixed some rounding errors to could cause frame rates when PerformanceTest was first starting to be displayed incorrectly
- Added support for some AMD 8700M/8800M video cards and updated System information library
V8.0.Build 1028 - 8th/January/2014
- Fixed a bug where in some cases the calculated overall 3D graphics mark was lower than it should have been, this mainly affected Intel integrated video cards.
This would only happen if the DX10 test failed to start but would start on a second instance, in these cases a penalty was applied as if it had no DX10 compatible video card, despite running the test (related to the changes made in v8.0 build 1024).
- Updated system information library: Added support for RAM modules on Supermicro server boards, SMBus support for Intel X79 Express Chipset, Fixed incorrect memory size reported for RAM modules > 32GB, AMD Athlon(tm) II X4 6xx socket type corrected.
V8.0.Build 1027 - 9th/December/2013
- Added support for AMD R9 290, 270 and R7 250 video cards
- Fixed a possible crash when retrieving hard disk information
V8.0.Build 1026 - 28th/October/2013
- Added support for AMD R9/R7 series video cards
- Added some extra detail to error messages when trying to run the 3D tests and monitor is in portrait mode
V8.0.Build 1025 - 10th/October/2013
- Fixed a crash that could occur when trying to start the advanced disk test when PerformanceTest had not finished collecting system information
- Updated system information library with numerous CPU updates: Ivy Bridge-E, Silvermont, Bay Trail, Cloverview, Richland, Kabini, Temash, Delhi, Seoul, Abu Dhabi, Kyoto, AMD Family 16h (Kabini).
- Added icon to desktop when installing
V8.0.Build 1024 - 27th/August/2013
- Fixed a bug where on systems with Intel HD4600 graphics the DX10 test might not start first time and skip without an error message displayed, though a second attempt would start correctly.
V8.0.Build 1023 - 20th/August/2013
- CPU information and temperature updates: Added support for Cloverview, Richland, Kabini, Temash, Delhi, Seoul, Abu Dhabi, Kyoto.
- Fixed a crash that could occur when trying to launch the DirectCompute test from the advanced menu when no compatible video card was available
- If the number of processes is changed in the preferences to be more than the default amount for the system then the test duration can now only be set to "very long" so score accuracy is increased
- Increased "very long" test time in preferences from 15 seconds to 20 seconds
V8.0.Build 1022 - 20th/June/2013
- Fixed a number of bugs in the system information gathering introduced in the previous release.
V8.0.Build 1021 - 14th/June/2013
- Added support for Haswell CPUs.
- Improved support for Atom CPUs.
- Added CPU turbo speed filter to advanced search.
- Fixed a bug causing the text export to print gibberish when the system information exceeded 512 characters.
V8.0.Build 1020 - 23rd/April/2013
- Fixed bug with Intel 220.127.116.1171 drivers (and possibly others) preventing loading of DirectX 9 DLLs causing PerformanceTest to incorrectly report that it has a corrupt install.
- Fixed bug where the CD test did not update the status window.
- Display details about total results found in basic search window.
- Disable wait icon after search in baseline search windows without user having to move mouse.
- Added baseline ID search to advanced search options.
- Fixed bug on some systems where extra RAM modules would be displayed.
V8.0.Build 1019 - 9th/April/2013
- Fixed crash when running the CPU tests on some systems.
- Fixed crash attempting to use /a automation command line parameters in trial version.
- Rotated video card model on main screen to point outwards from motherboard.
- Possbile bug fix for DirectX10 test on some Windows 8 systems.
V8.0.Build 1018 - 14th/March/2013
- Fixed crash bug introduced in build 1017 that affected 32-bit systems with Windows 7 or later.
V8.0.Build 1017 - 12th/March/2013
- Fixed bug preventing correct testing of more than 64 logical processors. In previous releases we relied on the O/S to distribute processes evenly between processor groups. It turns out this doesn't always work & we had to write code to explicitly assign processes to groups of CPUs to get a better load distrubution.
- Bug fixes for DirectX 10 test relating to failure to enter full screen mode. This warning message has been removed (after beening added in build 1016). But in it's place is a better (automatic) method of selecting the best monitor in a multi-GPU multi-monitor setup. We hope this will prevent some problems getting into full screen mode.
- Fixed flashing splash screen at startup.
V8.0.Build 1016 - 18th/February/2013
- Fixed crash when trying to save baseline.
- Stopped DirectX 11 test falling back to software rendering mode.
- Changed DirectX 10 test to abort if it failed to initialize fullscreen mode.
- Fixed a bug preventing the advanced disk test from doing raw access.
V8.0.Build 1015 - 8th/February/2013
- Fixed a bug incorrectly identifying bad system clock measurements.
- Fixed a bug causing a failure to save which drive was tested for later viewing in the graphs.
V8.0.Build 1014 - 6th/February/2013
- Fixed a bug causing the DirectX 10 test to crash when the primary desktop monitor was not monitor 1.
V8.0.Build 1013 - 25th/January/2013
- Memory latency test now also updated to run more reliably on system with limited memory.
- Added error case to DirectX 11 test to detect and warn user about erroneous results.
V8.0.Build 1012 - 18th/January/2013
- Fixed crash bug introduced with previous memory changes for systems with limited memory.
V8.0.Build 1011 - 17th/January/2013
- Fixed bug causing a crash when running a script using the /s command and the system did not support 3D, or the 3D was turned off using the /NO3D option.
- Made change to make memory tests more likely to succeed on systems with limited memory.
- Made change to threaded memory test to allow give it a greater chance to complete on very slow systems.
- Added warning at start up if system is running on battery power.
V8.0.Build 1010 - 27th/December/2012
- Fixed problem with rendering water in DirectX9 - Complex test on Intel cards.
- Added frame jitter measurement in advanced 3D test.
- Updated identification for some newer Radeon cards.
V8.0.Build 1009 - 11th/December/2012
- Changed DirectX9 tests to allow them to run on systems that don't support any anti-aliasing. Previously anti-aliasing was required to launch the test. Now the test will launch but a 25% penalty will be applied to the benchmark result to make up for the poor quality 3D rendering. This means the older Intel HD graphics CPU/GPUs can now run the DX9 test (e.g. the i3-530 CPU).
- Now show number of logical processors for all CPUs, rather than just Hyperthreading status for Intel CPUs. This will also mean that some AMD CPUs like the AMD FX-8120, 8350, 8150 will display 4 cores (and 2 logicals per core), rather than 8 cores.
V8.0.Build 1008 - 30th/November/2012
- Fixed bug introduced in build 1007 preventing information about the motherboard being correctly gathered.
- Improved detection of some AMD video cards.
- Fixed display bug in advanced search when 0 baselines were found.
V8.0.Build 1007 - 23rd/November/2012
- Fixed bug causing splash screen to remain open on some Windows 8 systems.
- Fixed bug misidentifying the hard drive tested on some systems preventing baseline from being uploaded.
- Removed temperature detection of AMD video cards on 32-bit systems as this was causing a crash.
- Fixed bug trying to load baseline files from previous versions using the 'File->Load Baseline' menu option.
- Reduced start-up time on some older CPU models.
V8.0.Build 1006 - 7th/November/2012
- Fixed DirectX9 tests to work with some older video cards (GeForce FX 5000 series and similar)
- Fixed bug causing current, min & max transfer rates in the advanced network test incorrectly showing average rather than instantaneous speeds.
- Advanced network test now shows the average speed at the end of the test, where the current speed was while the test was running.
- Fixed a crash when viewing disk results and certain disk related system information had failed to be gathered correctly.
V8.0.Build 1005 - 2nd/November/2012
- Added some extra command line parameters to selectively disable certain parts of the system information gathering.
- Fixed crash that occurred when trying to use the /s flag in shareware mode.
- Fixed a crash on some systems that occurred after running any of the tests.
- Improved startup performance slightly on some systems.
- Changed installer so that PerformanceTest doesn't automatically launch when installing in silent mode.
- Added server 2012 to list of OS options in advanced baseline search.
V8.0.Build 1004 - 22nd/October/2012
- Fixed expiry of certain features after 30 days even when using a licensed version of the software.
- Fixed bug with the USB install that prevented the DirectX 11 test from being run.
V8.0.Build 1003 - 18th/October/2012
- Added minimum and maximum speed display for Advanced Network Test.
- Fixed bug with the USB install that prevented the 3D tests from running from a USB drive.
- Fixed resource leak.
- Fixed crash in the Advanced Disk Test when using RAM drives.
- Added ability to disable temperature gathering from command line using SAFETEMP flag. (Can help if a crash is seen on start-up)
V8.0.Build 1002 - 12th/October/2012
- Fixed resource leak introduced with last release that caused instability after prolonged usage.
V8.0.Build 1001 - 12th/October/2012
- Fixed bug causing results to not display with certain Windows settings.
- Fixed crash at startup when a SoftPerfect RAM disk is active on the system.
- Fixed a bug that application instability after having opened the advanced CD/DVD window.
- Re-implemented SAFEMODE flag for V8, significantly reduces the amount of tasks performed at application startup.
- Fixed a caching issue with the new version check.
- Fixed a bug that caused scripting to freeze when reaching the DirectCompte tests.
V8.0.Build 1000 - 9th/October/2012
- New Single threaded CPU test and corresponding chart.
- New Direct 2D graphic test. Direct 2D is the relatively new hardware accelerated, immediate-mode, 2D graphics API from Microsoft. It requires Vista SP2 or above to work.
- New DirectX 11 3D test. To run this test a video card with DirectX 11.0 support is required, along with Vista SP2 or above as the O/S. This test exercises new video card features like Vertex and Pixel Shader 5.0 effects, Unordered transparency techniques (in the Jellyfish) and Tessellation technique to produce the detail in the terrain
- New DirectCompute benchmark test as part of standard GPU/3D tests. DirectX 11.0 video card support is required for some of the DirectCompute tests. Thanks to Jan Vlietinck & Keenan Crane who wrote some of the code we used in this test.
- New DirectCompute advanced test suite.
- New memory latency test, both as a standard test and advanced test.
- Re-written standard RAM tests in assembler to be less CPU dependent and more RAM dependent. We are also going to be making available some RAM performance charts
- Advanced tests for Blu-ray & DVD burning on Windows 7 and later.
- Ability to automatically run tests multiple times and take the max result.
- Improved advanced disk test. It is now possible to specify the level of randomness in the data, which can be useful for testing drives that compress the data.
- Scripting of the advanced disk test is now possible.
- Ability to run network test for a much longer period of time.
- IPv6 support in network test.
- Improved system information. Include RAM stick model and timing info, CPU temp & more.
- Re-balancing of instruction mix in the Integer maths, floating point maths and prime number tests.
- Encryption test now makes use of AES-NI CPU instructions where available.
- Faster start-up on launch despite additional system information collection.
- Better looking graphs / new charting engine in advanced tests. Can now graph multiple data series and customize the look of the graphs
- Support for testing more than 64 CPUs / cores.
- Better baseline search, both for basic and advanced search.
- Combining of baseline management windows, adding bookmarks, etc..
- User interface updates with 3D models.
- New baseline file format. We are now using the ".ptx" file name extension for PT8 baseline files. New format is effectively compressed YAML. It should be much more extensible than the older binary file format.
- New backend database structure for storing results. This was required as we expecting to get to around a million baseline results next year.
- Each uploaded result now gets a web page that displays the summary of the submitted results.
- Better video card model number detection, especially dual cards and Intel HD GPU.
- Removed the old bouncing balls 3D test as it was looking a bit lame.
- Modified the existing DirectX 9 and DirectX 10 3D tests to have better textures, use better models and improved shaders.
- Improved SLI & Crossfire support. SLI & Crossfire should now produce some performance improvements. We would have liked to do more here but nVidia and AMD were not able to provide any assistance.
- Official Windows 8 support.
- Official Windows Server 2012 beta support, with full support upon the Server 2012 release.
- Combining 32bit and 64bit releases into a single install package. The correct executable is autoselected at install time.
- Faster install after download by removal of some install dependences.
- More options for exporting, and copy / pasting the benchmark data into other applications, like Excel.
- As an experiment we are relaxing 30 day trial period, so that the core function can continue to be evaluated beyond 30 days. So rather than it just stop working after 30 days, the software will continue to run, with cut down functionality
- Bug fixes, of course.
V7.0.Build 1031 - 21st/August/2012
- Fixed an issue where some system information wasn't being correctly gathered on XP.
- Fixed an issue displaying the wrong OS in the baseline management window.
V7.0.Build 1030 - 7th/July/2012
- Fixed a bug in the 32-bit version of PerformanceTest preventing gathering of certain system information.
- Some minor changed/improvements to the advanced baseline search window.
V7.0.Build 1029 - 26th/April/2012
- Faster startup on some older CPU models.
- Improved support for Ivy-Bridge and Xeon E7 processors.
- Correct identification of Windows 8.
V7.0.Build 1028 - 23rd/March/2012
- Improvement for overclocking detection on some AMD Llano CPUs.
- Added correct identification of Radeon HD 7700 and 7800 series cards.
- Fixed possible crash when populating baseline information list.
V7.0.Build 1027 - 20th/February/2012
- Further improvements to CPU speed detection for AMD FX and 'A8/6/4' series.
- In the system information the display of turbo clock speed was changed from being measured clock speed in Mhz to theoretical max clock speed. Measured clock speed can be less than Max due to thermal issues.
V7.0.Build 1026 - 7th/February/2012
- Improved CPU detection on several newer AMD and Intel CPU models.
- Added detection for Radeon 7970 and 7950.
- Fixed a bug with detection of some older Radeon cards.
- Possible fix for an intermittent crash in the full screen planes test.
- Possible fix for crash bug in version check dialogue.
V7.0.Build 1025 - 24th/September/2011
- Fixed bug preventing some systems from running CPU tests.
V7.0.Build 1024 - 15th/August/2011
- Fix for flashing main window. Not confirmed to work with all system configurations.
V7.0.Build 1023 - 29th/July/2011
- Fixed a lockup when trying to do a disk test on a drive without enough space.
- Fixed a bug that would allow 2d results to go negative.
- Added warning when running 32-bit PT on 64-bit Windows
- Added AMD A series CPUs to the advanced baseline search.
V7.0.Build 1022 - 4/May/2011
- Minor improvements in detection of newer Intel CPUs.
- Now displays PT build and architecture in baseline information.
- Minor clarification about score weightings added to help file.
V7.0.Build 1021 - 10/February/2011
- Improved support of Intel Sandy Bridge processors.
- Fixed bug in advanced disk test that was causing the random seeking to not use the entirety of the test area in some cases.
V7.0.Build 1020 - 5/January/2011
- Fixed the visualized physics test image mode, which was broken.
- Improved CPU information gathering on VIA cpus.
- Improved CPU speed detection on overclocked Core 2 systems.
- Added detection for several Radeon graphics card types.
V7.0.Build 1019 - 27/October/2010
- Fixed a number of resource leaks that were causing PerformanceTest to become unstable after several hours of sustained usage.
- Improved a 3D test error message to give more information about the error.
- Added detection for Radeon 6850 and 6870 video cards.
V7.0.Build 1018 - 24/August/2010
- The string sort test had mistakenly been changed to be sorting strings 250 characters long instead of 25 characters. This occurred in build 1015 but went unnoticed until 1017. This change has been reverted. Any string tests run on versions 1015, 1016 or 1017 have degraded results.
- An issue preventing the CPU tests from running on Windows XP 64-bit has been resolved.
V7.0.Build 1017 - 18/August/2010
- Minor fix to SMBIOS data collection, can now collect data from a wider range of SMBIOS versions.
- Added detection for failed 2D tests where there was no display to render the test to. This has been seen when running PT via remote desktop and then closing the remote desktop window while the tests are running.
- Fixed crash on Windows 7 with advanced CD test when trying to write CDs. CD Write test is not supported on Vista/Win7.
- Improved CPU information gathering, in particular for AMD CPUs with Turbo Core technology.
- Increased timeout duration on CPU tests. The String Sorting test
can be slow to exit and on some lower end systems this
test would timeout resulting in no score being given.
- Fixed truncation of filename in baseline info window.
V7.0.Build 1016 - 30/July/2010
- Fixed 3D DX10 Test bug that could cause a crash on certain systems (seen on some systems with dual Xeon CPUs on 64bit Window).
- Pentium/Celeron M CPU speed detection improved.
- Refined memory limits on Visualised Physics Test and improved out of memory error case handling. Should be much harder to crash the application now when using this tool. As a result however the limits may now be lower on some systems, especially 32-bit systems.
- Fixed problem where copied system information was missing newlines when pasted into certain applications, in particular Notepad.
- Allowed use of Return key in the license key text entry field for entering new lines. However, it is still recommended that users copy and paste license keys rather than trying to type them manually.
V7.0.Build 1015 - 6/July/2010
- Improved Advanced Memory Test. The tests were optimized to be tighter loops and should now return higher bandwidth results, making the steps from level 1 to level 2 cache to main RAM much more obvious.
- Improved CPU Speed Detection on newer CPUs with turbo mode. Should provide a better indication of if a machine is overclocked than in previous builds.
- Fixed corrupt icon graphics in system tab when the software was running on high DPI displays.
- Fixed possible rare DX9 3D test crash in the "complex 3D test" were the test might crash after completion, during cleanup.
- There was a suspected out of memory crash fixed. We suspect some people forced a high number of CPU test threads on low end systems. The number of test threads is now limited by available RAM. 50MB of RAM is now required per test thread, (after the initial 2 threads).
- Added a function to collect mini-dumps in the case of crashes. This should make it easier to diagnose faults if people can send us the mini dumps.
V7.0.Build 1014 - 27/May/2010
- Fixed Advanced RAW low level disk write test. As the Advanced RAW test destroys the existing data on the disk, the test will now only work on unformatted drives. If you need to run this test on a formatted drive, you need to manually remove the partitions first.
- Fixed bug where the last data point of the advanced disk export would be printed twice
- Fixed minor text error in HTML export
- Added PT Version to HTML Export
- Fixed bug where PT would disable Turbo Mode on some Intel CPUs and not Re-enable it when done
- Fixed bug where new windows would sometimes open on wrong monitor
V7.0.Build 1013 - 15/March/2010
- Added i3 to list of searchable CPU types in advanced baseline search
- Fixed install to USB bug where CSS directory was not copied
- Fixed problem where user could launch multiple client network tests at once
- Fixed crash when trying to run advanced physics test without being able to render a 3D environment
- Improved SMBIOS information gathering
V7.0.Build 1012 14/January/2010
- Fixed bad reporting with more than 10 ram modules, only the first 10 modules will now be reported
- Fixed incorrect value being exported for random string sorting in advanced multiple process test
- Fixed loading of Pre-V6 baselines in the 64-bit version of PerformanceTest
- Fixed blue screen crashes on certain CPUs when running Windows Server 2008
- Fixed crash on start up on some systems with a large number of RAM modules (> 10 modules populated in SMBIOS)
- Better support for detection of CPU speed on new 32nm i3 and i5 CPUs with turbo mode
V7.0.Build 1011 10/November/2009
- Fixed a blue screen crash that affected certain CPU models
V7.0.Build 1010 29/October/2009
- Fixed bug showing Windows 7 baselines as Vista baselines in baseline selection window
- Allowed full results to be generated on DX10 systems without 4x AA
- Fixed program crashing when loading pre version 6 baselines
- Improved identification of some newer ATI Graphics Cards
- Added description of /p option to help file
V7.0.Build 1009 6/October/2009
- Improved CPU Information gathering, in particular for the core i5
- Also improved CPU speed detection
- Fixed issue in Win7 when using multiple monitors and running the 3D tests
- Added i5 to list of CPUs in advanced baseline search window
- Updated help file with better CPU test descriptions and correct 2D scripting commands
V7.0.Build 1008 9/September/2009
- Better core detection for i5 Processor
- Better DDR3 detection
- Better warning message for a known issue with complex full screen 3D test on Win7 with dual monitors
V7.0.Build 1006 6/July/2009
- Included Help files when installing to USB
- Better CPU clock speed measurement
- Fixed bug loading baselines in local folder twice when using /p
- Changed error 5 from error to warning in DX10 test on Win7
- Fixed a few misleading error messages
V7.0.Build 1006 12/May/2009
- Fixed "Find Same CPU" not working for certain CPU types.
- Stopped welcome window reopening at certain points when running in evaluation mode.
V7.0.Build 1005 16/April/2009
- Fixed DX10 test to work on videocards that do not support 8x AA
- Changed Passmark Rating calculation so that CD scores have a smaller impact and can only ever increase the overall score
V7.0.Build 1004 26/March/2009
- Fixed a bug that allowed uploading of incomplete baselines.
- Fixed a bug that was causing some types of old AMD CPUs to get extremely high scores.
V7.0.Build 1003 19/March/2009
- Fixed a possible crash in the Medium/Complex 3D test.
- Fixed an error in reporting the transfer speed in the advanced network test.
V7.0.Build 1002 4/March/2009
- Fixed an error when running licensed from USB Drive.
V7.0.Build 1001 4/March/2009
- Fixed crash on some systems when opening "print" menu option.
- Fixed baselines info not displaying when selecting for display from the web download interface.
- Fixed OS lacking 32/64-bit info in baseline search.
- Fixed problem with installing a licensed copy of PT to a USB drive.
V7.0.Build 1000 25/February/2009
- New DirectX 10 3D Test (Requires Windows Vista and a DirectX 10 compatible video card)
- Large improvements to the user interface, including near completely reworked main window.
- Can now view the system information from multiple baselines simultaneously.
- streamlined baseline selection dialogue
- Increased the default size of the main window to go along with the graphical improvements.
- Improved baseline search functionality.
- Removed almost all restrictions in the trial version.
- Increased the amount of system information gathered and displayed.
- Improved copy options for quickly getting system information from PT to the clipboard.
- /p option added to use the application directory instead of the user directory
- Added EXPORTBASELINE script option to export .pt file
- Overhauled most of the 2D test suite. Complex Vector tests and Image filtering tests now included.
- Changed fonts and text test to have text which is more representative of a real document.
- Increased intensity of "Windows Interface" test.
- Increased the resolution at which 2d and 3d graphics tests are run to place more load on the video card.
- Prime numbers test optimised.
- String sorting test intensified to use more memory.
- Image rotation test was moved from the CPU suite and added as part of the image filters in the 2D suite to show what is happening.
- Added CPU Physics simulation test
- Added Visualised Physics Test to advanced test suite. This is a customizable 3D representation of what the new physics CPU test is doing.
- Increased unchached memory test from 4MB to 64MB as many CPU's have 4MB caches now.
- Updated image export functionality.
- Install to USB drive menu function.
- Changed Passmark Rating calculation to limit the effect of a single extremely fast system component.
V6.1.Build 1018 22/September/2008
- Correction to a rare crash bug when saving a report or doing a
HDD benchmark (when a physical disk is numbered greater
- ATOM CPU Hyperthreading and number of CPU's detection added.
- CPU model number changed to include extended model information for newer CPU's.
- CPU L2 cache sizes for newer Intel CPU models added.
- Debug logging increased for CPU detection.
- A startup problem for an unreleased CPU corrected.
V6.1.Build 1017 12/May/2008
- Removed some debug information to better reflect performance in particular Virtual environments.
V6.1.Build 1016 1/May/2008
- Improved the disk tests for compressed disk volumes.
- Improved the CPU frequency measurement introduced in PerformanceTest v6.1.1013 for certain newer Intel CPU's.
- Corrected a bug where PerformanceTest V4.0 baseline files could not be loaded into PerformanceTest v6.1.
- Corrected a crash bug on startup for a rare BIOS configuration.
V6.1.Build 1015 16/Apr/2008
- Corrected a crash bug on startup for a rare BIOS configuration.
V6.1.Build 1014 2/Apr/2008
- Corrected a bug introduced in v6.1.1013 where the default number of
CPU test processes was 1, rather than the number of CPU's. This
resulted in much lower CPU scores if this default was not changed
V6.1.Build 1013 27/Mar/2008
- Changed the startup of PerformanceTest to better support Virtual environments like VMWare.
- Changed the mechanism to calculate the current CPU frequency as newer Intel CPU's (e.g. Core 2 Duo) that have been overclocked by lowered their bus speed multiplier may reflect a speed based on a higher bus speed multiplier. This does not impact the CPU benchmark scores.
V6.1.Build 1012 26/Feb/2008
- Implemented a work around for the error "the procedure entry point IsUserAnAdmin could not be located in the dynamic link library shell32.dll". This occurred on systems with old versions of shell32.dll.
V6.1.Build 1011 28/Jan/2008
- Added searching for online baselines for AMD Phenom and Intel Core 2 Extreme.
- Removed the restriction of only allowing a single instance of PerformanceTest to run at one time. This allows multiple Advanced Network tests to be run concurrently.
- Minor changes to license management.
V6.1.Build 1010 4/Dec/2007
- Corrected the Multitasking test to run across multiple CPU cores.
V6.1.Build 1008 22/Aug/2007
- Corrected a software failure bug on startup on a few systems where the system OEM had not populated Baseboard information.
- Corrected a software failure bug on startup (particularly Vista) where a DirectX function was causing software failures in "dsetup.dll".
- In some cases where the system OEM has not populated Baseboard information, the display in the Baseline information window could be gibberish. This has been corrected.
- The Fixed block size in the Advanced Network test is now used rather than the maximum block size.
- The Intel Core 2 Quad has been added to the search function for downloading Baselines from the Web.
- Quad core L2 cache now detected.
- Included an option to skip the DirectX warning message on startup, either by specifying the existing command line parameter of "/i" or by selecting "No" to the DirectX warning the first time it is displayed.
- Included a Vista specific warning message for Vista DirectX10 to help avoid confusion over the DirectX9.0c requirement for PerformanceTest.
- In very, very rare cases, the CPU clock speed could be measured and the result a clearly ridiculous value. This has been corrected.
- Minor correction to the HTML export to better handle errors opening the output file.
- The filename displayed in the error message when the exported text report file can not be opened has been corrected.
V6.1.Build 1003 1/May/2007
- Added the ability to throttle the UDP data throughput with the Advanced Network test.
- Minor bug corrections.
V6.1.Build 1002 3/April/2007
- Resolved an issue where PerformanceTest would fail to start on systems with DEP enabled for all programs.
- Corrected the Start menu help link.
- Minor bug corrections
- Ceased support for Windows 98 and ME. Please see our website for an older
version supporting Windows 98 and ME.
- Corrected a bug where the number of cores reported on a Quad core system was incorrectly reported as CPU packages.
- Improvement in Graphical user interface.
V6.1.Build 1001 28/December/2006
New point release
- Support for Windows Vista.
- Previous versions of PerformanceTest required access to a range of resources that, under Vista, were only accessible when PerformanceTest was run with elevated privileges. The following changes have been made to address this. The changes have also been made to Windows 2000 and XP for a consistent user experience.
- Windows 2000, XP and Vista:
- Files have been moved from the Program Files directory to the User Personal directory. Examples: Vista - "C:\Users\<username>\Documents\PassMark\PerformanceTest", XP - "My Documents\PassMark\PerformanceTest\". These include: The baseline installation directory, Download Baselines from Web, Upload baseline, all HTML report default directories, default Save As Baseline and image, default Export results and Export Results For Web, Debug logging and the CD burn test image.
- PerformanceTest will now check both the Program Files and User Personal directory for the following files: a key.dat file (for running PerformanceTest from CD or USB drive), for installing baseline files and for installing Cascading Style Sheets for HTML report formatting.
- Files have been moved from the Program Files directory to the User Application directory. Examples: Vista - "C:\Users\<username>\AppData\Local\PassMark\PerformanceTest", XP - "C:\Documents and Settings\<username>\Local Settings\Application Data\PassMark\PerformanceTest". These include: The configuration file and temporary files such as statistics files for the Advanced tests.
- For disk tests (standard, advanced and multiprocess), an option has been added to allow the User Application directory to be used for the test (as the user may not have access to any of the root directories to perform the test).
- If the User is running Vista and is not running PerformanceTest with elevated administrator, some system management BIOS information will not be available (such as number of RAM slots). A button, "Get info", has been added to the Baseline information window to run a program to collect this information. The user will be requested to accept elevated administration for this operation.
- The previous version of PerformanceTest did not collect the Video Card
description on Vista. This has been corrected.
- All operating systems:
- DirectX Direct3D has been updated to 9.0c October 2006. This was mainly to correct a bug in DirectX where temporary texture files were written to the root directory, which in most cases the user did not have access to when using Vista. PerformanceTest now includes the update installation for DirectX Direct3D October 2006.
- The Baseline selection Window now displays the Operating System for Vista and Longhorn correctly.
- The error message "This program might not have installed correctly" after installation is no longer incorrectly displayed.
- To provide the ability to automate quite complex testing scenarios, scripting of Standard preference configuration settings and test selection has been added.
- To assist test automation, an option to suppress the main test warning
and error messages has been added to preferences.
- Increased the maximum number of CPU processors that can be tested
from 10 to 64.
- Added a new formatted text result report.
- Browser based HTML help replaces the older style of Windows help.
- The Baseline selection Window is now resizable.
- The CPU type, rather than the manufacturer is now displayed in the Baseline selection Window.
- Saved results are now all shown to 1 decimal place, and RAM is shown in MegaBytes rather than Bytes.
- The 64-bit version of PerformanceTest better reflects current 64-bit application performance when compared to the previous versions of PerformanceTest, which reflected early 64-bit applications.
- The set of baselines provided with PerformanceTest provide a wider spread of PC hardware technologies.
- Additional system information is now available on 64-bit Vista systems as the Device driver PerformanceTest uses is now digitally signed.
- The User Interface graphics and icons have been updated.
- Bug corrections:
- During the 3D test, an error Window could have been obscured by the main PerformanceTest Window. The user may have thought that PerformanceTest had stopped responding.
- On the result graphs, a long CPU type description could continue off the Window. The CPU type is now split across 2 lines if required.
- "Install Baseline from Web" would fail if "Install Baseline from File" had been used beforehand.
- The Advanced Disk test could display the status "stopping", instead of error in an error case.
- If RAM slot information was not available, the device information for the memory test would show 0 of 0 slots.
- The PerformanceTest Version number inconsistently used "V6.0" and "6.0". The V is now ommitted.
- Some additional checks to ensure corrupt OEM populated SMBIOS data does not cause PerformanceTest to close on startup.
- The title bar icon for the Download from Web Window has been corrected.
- Corrected a bug where-by on rare occasions PerformanceTest could close when selecting Next or Previous on the Download from Web Window.
- The Download from Web Window error messages must be OK'ed before continuing.
- The main window is now updated to reflect any changes made in preferences.
- In all of the Save As Windows, a blank filename was accepted and saved as e.g. ".pt", this is no longer accepted.
- Selecting a Graphics file name to save would incorrectly set this as the default Baseline name to save.
- PerformanceTest would close when in the Save as Test Window, the filename was blank and a different type of report was selected. This has been corrected.
V6.0.Build 1013 1/September/2006
- Corrected a bug in V6.0 1012 where the CPU test would not run if the
system did not have the latest Microsoft Runtime library msvcr08.dll (this
should not have been required).
V6.0.Build 1012 29/August/2006
- Changes to the CPU tests to overcome a problem on some dual core systems where PerformanceTest could crash. This problem was reported mainly for AMD X2 systems. Error displayed: "access violation at 0x7c91084b".
- Intel Core 2 and Intel Core Duo have been added to the Online Baseline search.
- Correction to the uploading of a Baseline file where the filename is specified by the user (i.e. not the current results).
- A baseline with a result of 0 can no longer be uploaded.
- On startup in x64 versions of Windows, a PerformanceTest message window is obscured by the PerformanceTest splash screen. This has been corrected.
- HTML help has been added for Windows Vista and Longhorn Server.
V6.0.Build 1011 28/June/2006
- Improvements to the Advanced disk test:
(i) Reduced the minimum file size from 1MB to 128Bytes.
(ii) Allow partial results to be graphed and exported.
V6.0.Build 1010 19/April/2006
- The Large RAM test could have failed on a system with a large amount
of RAM that was fragmented. PerformanceTest will now continually attempt
to allocate successively smaller amounts of memory and hence avoids this
- Under Windows 2000, XP and 2003 server, the system data (SMBIOS) was unavailable in rare circumstances. This has been corrected.
- Searching for "Pentium D" and "TransMeta" CPUs
has been added to the Install baseline from Web.
- When the OEM has not filled in the recommended CPU string, PerformanceTest now fills the string in with a little more detail.
V6.0.Build 1009 24/February/2006
- Corrected a bug where the 3D Terrain test could crash in build 1008.
- Under certain circumstances the number of CPU cores was not reported in the HTML report. This has been corrected.
- On systems with a very large amount of RAM or highly fragmented RAM, the large memory test would fail with no error message. This test now adjusts the test memory for this situation and reports an appropriate error if the problem still occurs.
- When running PerformanceTest from CD, the Advanced CD test failed if the PerformanceTest CD was the CD under test.
- When running PerformanceTest from CD, the Advanced test graphing function is not available, however a blank graph was just shown. An information message is now displayed.
- The Advanced CD test could under certain circumstances report that it was 99% complete, when it was 100% complete. This has been corrected.
- Corrected a bug with the Advanced Disk test graph, where the disk size was reported as 0MB for disks smaller then 1GB.
- Some Pentiums (where the CPU does not support the reporting of the number of cores - this can still be a single core CPU), PerformanceTest indicates it can not detect whether Hyperthreading is supported.
- Added a specific error message if the 3D Terrain test is run from a
read only location (e.g. a CD) and DirectX fails to create the terrain
V6.0.Build 1008 rev.2 23/January/2006
WIN32 build 23 / January / 2006
- Change to try to resolve a startup issue.
V6.0.Build 1008 18/November/2005
WIN32 build 18 / November / 2005 and WIN64 build 22 / November / 2005
- Corrected a bug where a long Baseline download filename was not displayed on the Baseline download Window (or possibly closed PerformanceTest).
- If the number of CPU test threads was set to 0 (by the user or a
CPU information detection problem) then the CPU tests appear to run
(but don't) and no CPU results are shown. The number of CPU test threads
must now be between 1 and 10
(inclusive), with a default of 1.
- In the case where PerformanceTest does not have access to the number of logical processes supported by the Operating system, PT may have exited without warning on startup. eg. systems running VMWare. PerformanceTest will now run, but with an indication that Hyperthreading could not be detected.
- Test results have been made more consistent across multiple test runs.
- The 3D test now invalidates the 3D test results and displays an error or warning if (i) the user presses Escape, (ii) a Popup removes focus from the 3D test window (iii) there is a rendering error.
- To improve accuracy of low results displayed, for results less than five, two decimal places will now be shown.
- Advanced Memory: Capped the test memory to 0.5GB, as test on systems with a large amount of RAM may fail due to memory fragmentation.
- Corrected a bug where this error condition was not correctly handled, and the "Exit" button was inaccessible.
- Corrected the Advanced Network Test "Connection setup failed" status message.
- Advanced 3D: Increased the size of the Adapter combo and truncated the string if too long.
- Advanced CD: Test file directory control expanded.
- Save As baseline: Window size increased to fit Number of CPUs
- Bug entering licence key has been corrected.
V6.0.Build 1006 16/September/2005
- Improvements to the Advanced Network test. Includes the addition of reporting the number of UDP test data blocks lost per reporting time period.
- Corrected the number of CPU cores reported for AMD Dual core CPU's. Previously CPU cores were reported as CPU packages.
V6.0.Build 1005 5/September/2005
WIN32 and WIN64 builds
- A new "game like" 3D test has been included as a standard
and an advanced test. This test will make use of some of the advanced
features in new graphics cards.
- To make it easier to compare your results with similar types of computers,
Baseline searching, downloading and uploading is now possible from with-in
- A new Memory test has been added that measures the performance of the memory sub-system for applications that use a large amount of RAM.
- The Advanced Memory test now allows the testing of 64-bit data (in
addition to 8, 16, and 32-bit data).
- A new Advanced CD test to measure the performance of CD-R/CD-RW writing
has been added (under Windows XP and 2003 server).
- The Export for Web feature allows the comparison between the current computer and all selected baseline computers to be exported in HTML. This has been improved to highlight the highest scoring test results for each test, test category and overall result.
- Changes to CPU test:
- A new CPU test has been added - Prime number test;
- The MMX test has been removed;
- The CPU - Integer Math's, CPU - Floating Point, and all memory tests, that now test with a combination of 32 and 64-bit values (rather than just 32-bit values as was the case in PT5.0).
- The standard disk tests have been changed to improve the accuracy
and consistency of results, particularly for high end hard disk drives.
- Test re-weighting. Due to advances of technology components at different
rates, re-balancing the results between major tests has been undertaken
to achieve our intended importance/weighting of each test (as described
in the help file).
- The Graphical User Interface has been updated.
V5.0.Build 1036 30/June/2005
- WIN64 specific: Due to a change in defaults from DirectX 8 to 9, the Windowed 3D tests (Simple and Medium) may not have run as fast as possible, and hence yielded lower scores. PerformanceTest has been changed to measure the maximum 3D performance under DirectX9.
V5.0.Build 1035 15/April/2005
WIN32 and WIN64 builds
- Corrected a bug introduced in build 1033 with the SSE test.
V5.0.Build 1033 11/April/2005
WIN32 and WIN64 builds
- The Advanced Memory test now includes a Qword (64-bit data element)
test data size. This is useful in showing the benefits of 64-bit memory
access with a 64-bit system.
- The WIN32 version of PerformanceTest can now read a WIN64 produced baseline for comparison). Previously only the WIN64 version could read a WIN32 produced baseline.
- In the Baseline list, Windows 2003 Server is now also displayed for
the Operating system.
- When printing the results from the Menu "Print", any "Device
info" was printed so small it was unreadable. This has been corrected.
- In the Advanced Network test, the export results no longer prints
the column of unused "number of packets sent/received".
- The help file has been modified to further clarify the meaning of
the comparative Results.
V5.0.Build 1032 25/February/2005
- Corrected a bug where the display of some memory sizes was not correctly displayed with systems with more than 2GB of RAM. Note that the WIN32 version of PerformanceTest reports a maximum of 4GB of memory.
- To help interpretation of the Disk test scores in the baselines, some additional information has been added to the "Device Info" to indicate whether the disk is Removable or a RAM disk. This will help baseline users determine whether a very high disk score is the result of using a RAM disk or a removable drive like a USB 2.0 memory card/stick. This information will be available for baselines created with V5.0 1032 or above.
- In certain circumstances, systems with a large amount of memory (e.g. 3GB) have reported very high disk scores that do not strictly reflect the performance of the Disk I/O subsystem. The disk test has been improved to report the Disk I/O subsystem performance for these types of systems.
V5.0.Build 1031 16/February/2005
- Added the command line option SAFEMODE to start PerformanceTest without trying to obtain SMBIOS information. This may help with problems starting PerformanceTest, due to corrupt SMBIOS or SMBIOS not configured by the OEM. eg. PT.exe SAFEMODE.
V5.0.Build 1030 29/January/2005
- PerformanceTest crashed on startup if the System Management BIOS information pointed to data at invalid physical addresses. A sanity check for valid SMBIOS data addresses has been added and an warning displayed by PerformanceTest if this is found.
V5.0.Build 1029 17/January/2005
- Added debug information for a reported SMBIOS crash in W98.
V5.0.Build 1028 12/December/2005
- First Windows64 bit version of PerformanceTest. Runs under a 64bit version of Windows natively.
- WIN32 MMX and SSE/3DNow! benchmarks have been replaced by 64-bit Integer
and 64bit Floating Point tests, respectively.
- To allow easy comparison of baseline results, WIN32 baseline files
are compatible with WIN64 baselines.
V5.0.Build 1027 10/October/2004
- Increased the Advanced disk test IO queue length for asynchronous tests to 128.
V5.0.Build 1026 23/August/2004
- Minor User interface improvements.
- Increased DEBUGMODE reporting for SMBIOS on application startup. Check added for invalid SMBIOS table termination.
V5.0.Build 1017 30/July/2004
- Advanced disk test now checks the total disk test file size across all test threads before starting any test.
- Unformatted key drive handling.
- 2D graphics test now produces linear results based on test duration.
- Graphs now display 10 different colors for each test thread (instead of 5).
- Windows 98/ME corrections to:
- Unable to complete Advanced Disk Test.
- File->Export results for Web option does not allow choosing of a cascading style sheet.
- Baseline Information option doesn't show Disk information.
- Tooltips do not work on top menu bar
- When running 2D -> "Fonts and Text" Test under Win98SE, an error box will pop up and the user cannot click on the "Yes" or "No button.
- Installation now warns of existing directory.
- Advanced 3D test error messages are no longer 'Hidden'.
- A specific error message has been added when a user tries to install an already installed baseline.
- It is now indicated in the Baseline information window whether onboard device and system slot information is unavailable.
- When a PC comes out of Sleep mode, under all certain circumstances the status bar is now correctly redrawn.
- Average WAW in the advanced disk test has been corrected.
- Viewing a graph when there are valid results for 10 or more threads is now correctly handled.
- Menu item ellipses (...) have been updated.
- The setting box in the advanced 3D options no longer becomes incorrectly ungrayed.
- When the user stops the Advanced disk test, the progress bar is immediately set back to the start, rather than when the test is next started.
- The cascading style sheet is now correctly applied when exporting a "file to the web" to all directories, not just the installation directory.
- The "Office" cascading style sheet has been removed, as it was the same as the "Default" cascading style sheet.
- When the advanced disk test is stopped by the user, the status now indicates stopped rather than completed.
- The PerformanceTest folder is now removed in all cases after uninstalling the program.
- The existing testing utilizing Hyperthreading enabled computers improved.
- Rounding precision improved in the display of free disk space on
the with baseline information window.
V5.0.1014 Release on the 24/June/2004
- Corrected the handling of stopping an advanced disk test during a test file create.
- Corrected the initial display value for Ave. WAR, WAW, RAW and RAR.
- Included detection of old versions of riched20.dll for the 2D Fonts and scrolling test.
- Updated help file included.
V5.0.1013 Release on the 15/Apr/2004
- Corrected handling of an SMBIOS v3.2+ issue that has been found on
a very small number of PCs, that caused PerformanceTest to crash on
V5.0.1011 Release on the 19/Mar/2004
- Bug correction - collection of disk model number "unknown" for all disks, when one or more non-SMART drives connected. Device information displayed per disk, rather than being common across the 3 disk tests. Disk device information now changed to reflect changes in disk preferences.
V5.0.1010 Release on the 17/Feb/2004
- Advanced network test optimized for Gigabit network testing.
V5.0.1008 Release on the 9/Jan/2004
- Bugfix for saving disk baseline information
- Correction of rare startup crash bug
- Bugfix related to baseline hyperthreading information
- Advanced tests no longer report CPU load of 0.0% when it is unable to determine the CPU load on that specific system.
V5.0.1000 Release on the 10/Dec/2003
- Whole new visual look for the main results window
- Replacement of the Maths test suite with a more general CPU test suite.
This new CPU test suite also has better performance with multiple CPU
and Hyper-threaded machines.
- New 2D Graphics test: Fonts and Text. Tests typical rendering of text and embedded document images by scrolling, zooming, and font changing.
- New 2D Graphics test: GUI. Tests the rendering and interaction of Windows GUI controls such as list boxes, progress bars, tree view, and window movements.
- New CPU test suite, with better performance for multiple CPU and hyper-threaded machines.
- Advanced multi-process test: allows you to run multiple test processes at the same time to test your CPU's multi-tasking capabilities.
- Advanced disk test: new test window, allows you to specify and run multiple disk test threads. Also, additional information and statistics reported.
- Advanced CD/DVD-ROM test now identifies DVDs and displays the appropriate "times" factor relative to read speed (eg. "2x DVD")
- Updated decoding of level 2 cache display to handle newer CPU's such as the Celeron 2.2Ghz.
- Additional statistics for the advanced disk test.
- Merging of the MMX/SSE tests and the CPU tests and the rewriting of the MMX/SSE tests to be more realistic.
- Hyperthreading is now taken into account when displaying the number of CPUs installed.
- UDP protocol network testing, in addition to TCP
- Bug fix: PT was crashing when SMBIOS data was not found. This is
now handled correctly
- Main test results can now be exported for the Web with customizable stylesheets.
- HTML Web Reports supports user defined stylesheets, which can be
selected from a drop-down box in the HTML Report Options dialog.
- Additional system information is now displayed in the baseline information window. Including Motherboard information, PCI slot information, and memory slot information.
- New baseline file format to store the additional system information.
- The baseline information window was given a new look with scroll bars
to handle more text
- DEBUGMODE command line startup option
- Increased maximum packet size for network test to 16385 from 4096
- Hardware component display
V4.0 Released on the 1/Sept/2002
- Advanced disk testing with graphing,exporting and CPU utilization.
- Advanced CD/DVD testing with graphing, exporting and CPU utilization.
- Advanced 3D graphics testing for full screen and windowed modes
- Advanced network bandwidth testing with graphing exporting and CPU utilization.
- Advanced memory bandwidth testing and step size testing with graphing and exporting.
- Replacement of the standard 3D graphics test using DirectX 3 with three new DirectX 8 tests.
- A new standard test for CD / DVD drive benchmarking has been introduced.
- Much of the internal code has been rewritten to support foreign languages (French, German, Japanese).
- Many other minor improvements.
V3.5.1002 Released on the 12/Feb/2002
- HTML Export function now exports system information for logging and inventory purposes.
- Bug fix to correct Level 2 cache display for Pentium 4 CPU's with
more 512KB cache.
V3.5.1000 Released on the 29/Oct/2001
Minor release to support Windows XP
- Improvement to the user interface to better match the Windows XP look and feel. (Only available when running XP)
- Identification of the XP operating system in the system information window
- Correct Identification of video cards under XP
V3.4.1002 Released on the 3/Oct/2001
Minor release to fix a few bugs.
- Bug fix. A rare crash occurred while saving a baseline file.
- Correct reporting of RAM in machines with > 2GB of RAM.
- Bug fix. Baseline files with path names longer than 100 characters could not be loaded
V3.4.1001 Released on the 3/Sept/2001
Minor release for test automation.
- New command line parameters for automated testing from a batch file. See the help file for more details.
- Bug fix to MegaFlops test, which returned 0 on a few rare PCs with very low resolution timers.
- Bug fix to CPU type detection. Some early model Pentium II 300Mhz were reported as being Celerons.
V3.4 Released on the 21/Apr/2001
Minor release for new CPU types.
- Work around for bug in AMD Duron processor that caused L3 cache to be reported as 1KB. See AMD technical document, "TN13 revision 3" for more details of the bug.
- Detection of Intel P4 CPU and new AMD chips.
- Baseline version number is displayed in System info and baseline list.
- New SSE and 3DNow! test added in MMX test suite.
- Correction of a minor display bug in the baseline selection window for CPUs faster than 1000Mhz.
- New command line parameters for automation of the testing process (eg /ah autoruntest.htm).
V3.3 Released on the 12/Nov/2000
Minor release to implement disk test graphing.
- Disk speed test graphs have been implemented. Both read and write speed can be graphed in detail, with options to print the results, copy the results & adjust the X-Axis units.
- Improved overhead calibration method for MegaFlops test results should now vary less between runs.
- Statistics are now collected for the disk read and
write tests. They are placed in four files in tab separated format:
- Extra delay inserted before Read test to avoid overlap with cached test file creation. This can have the effect of making the read test result look slightly better.
V3.2 Released on the 12/Oct/2000
Minor release to improve disk testing.
- Disk cluster size is now displayed in the system information dialog and stored in all new baselines files.
- File system (eg. FAT32) is now displayed in the system information dialog and stored in all new baselines files.
- The disk test has been altered to always use 16KB blocks for read and write operations. Previously the cluster size for the disk being tested was used. (This may cause the results for the disk tests to vary compared to the results obtained in Version 3.1 & 3.0).
- Fixed bug in display of disk size in the system information dialog. Previously the size of Drive c: was always displayed, even if D: was being tested.
- The AMD Duron CPU is now identified.
- Windows ME is now identified.
V3.1 Released on the 1/Sept/2000
Minor release to correct some serious bugs.
- A problem in the calculation of the overall PassMark rating was discovered and corrected. The results from the 3D test and MMX tests were not being fully taken into account. This means that the release 3.1 PassMark rating can NOT BE DIRECTLY COMPARED to the 3.0 rating. Baselines saved in versions 1,2 & 3.0 are automatically re-scaled, upwards, when loaded into version 3.1 however. (so all previous baseline files can still be used in 3.1).
- Fixed spelling of 'Subtraction' for MMX tests.
- Fixed integer overflow error causing negative results when long duration is set with fast CPUs.
V3.0 Released on the 15/Aug/2000
- Better CPU speed measurement. In some configurations, such as multiple CPU systems, V2.0 was reporting the wrong speed. The test is now much more accurate on all systems. In some cases we can even measure to 1/10th of a Mhz.
- Improved differentiation of similar CPU types. Previously it was not always possible to detect the difference between some Celerons and Pentium processors.
- New logo was created for the Welcome screen and the about window. The new logo keeps a common theme with our other products.
- Added a splash screen.
- Changed the Operating version text from "NT5" to "Windows 2000".
- Dual CPUs are now supported in the baseline info window.
- Video display hardware is now detected, displayed and saved with baseline files.
- Added conversion routines for reading baseline files from previous versions.
- Added a drop down list for selecting the disk drive letter in the preferences window.
- It is now possible to give a name to the computer being tested from the preferences window. This is useful when publishing or exporting results.
- Reformatted the way the main window was laid out. The previous design was confusing some novice users. The % column heading was removed, a scale was included with check marks and extended units of measurement text is included above the scale if the window is wide enough.
- The bar chart can now be displayed in Multicolour or in two colors.
- A new test has been created. The test measures maximum number of floating point instructions that can be executed per second (FLOPS). The test is multithreaded (unlike existing tests) and will make full use of multiple CPUs if available. The test result also has zero overhead. The results of this industry standard test are also included in the summary window.
- A progress bar and the name of the current test is displayed when there is more than a single test being run.
- The scaling of the bar chart is now more progressive. The maximum value will now increase in smaller increments.
- Redid buttons, moved from large buttons to small buttons and added 3 extra buttons.
- Included MegaFLOPS in the Summary window.
- Updated and improved online help file.
- Added help buttons to a number of the dialog boxes.
- Redid the auto-installer to have the option of starting PerformanceTest immediately after installation. The size overhead of the installation program is now much smaller as well.
- The .pt file name extension is registered with Windows. This means you can double click on a ".pt" file in explorer to see the baseline results.
- Implemented scrolling with the mouse wheel.
V2.0a Released on the 23/Jan/2000
Minor bug fixes to correct two significant issues.
There seems to be a bug in DirectX that causes the textures required for the 3D tests not to load on a very small number of computers (this fault may even be associated with a particular video card / driver). V2.0a has a patch to work around this problem. The error message associated with this problem is, "..Error loading the required texture Bitmap file from disk..".
With a small minority of video cards, the unfilled shapes test doesn't run for the correct amount of time and the result for this test is too low. This is often associated with the filled rectangles (bitmaps) test returning a result that is too high. The problem is related to the speed of the 2D hardware acceleration and the CPU speed. The problem can occur if the CPU can request the display of images a lot faster than the video card can display them. V2.0a corrects this problem.
A number of error messages were also improved.
V2.0 Released on the 1/Nov/99
Major release with a lot of improvements over version 1.0.
- A button to run all the tests sequentially was added.
- When several disk tests are run, the test file is not recreated for each test. This means that the tests can be completed quicker.
- The disk tests now read and write information in larger blocks. In version 1.0, 1KB blocks were used. In this version the cluster size is determined and all disk reads and writes are done a cluster at a time. The cluster size varies from computer to computer. The average is around 32KB. This makes the test closer to the real life. Transfer rates are still reported in MB/Sec.
- The disk tests were redone. There were about 20-30%
of computers that did strange things when the Windows cache was disabled.
Some ignored the command, some went to far and disabled the disk controller
HW cache, etc..
The results were all over the place in some cases; so the test was dropped. We now use the cache but overload it so that we see the performance of the disk behind the cache.
- An additional Disk Seek plus read and write test was added.
- It's now possible to export the results in several "text" formats. This includes HTML, Comma separated values, and tab separated values.
- Added a registration button on the tool bar.
- Extra information is now saved with each baseline. This includes Drive size, Driver letter (eg. C:, D: ).
- Some code was reworked to improve compatibility on early Windows95 machines. We have even run PerformanceTest on some old 486 machines using the original Windows95 (+Directx).
- It's now possible to view baseline results without re-running all the tests.
- Multiple Baselines can now be viewed on the screen at one time. Up to 20 computers can be compared to each other at a time.
- The Select Baseline window was introduced to managed the addition and removal of baselines from the screen and display information about all the baselines available.
- The graphics on the display was improved and graphs now re-size themselves to the window size.
- The size of the notes field (in the Save as Baseline window) was increased to allow the user to add more information.
- Some tests display information about their environment. The graphics tests display the color depth next to each baseline. The disk tests display the drive letter.
- A problem with the Microsoft C++ compiler that caused inconsistent results for the MMX tests was worked around. The problem was to do with the alignment of 64bit numbers on the stack and caused some computers to report low MMX results.
- The printing was improved to handle the differences between landscape and portrait. A page header and page numbers are also included.
- The online help was updated and improved.
- Scrolling is now smoother. The Home, End, Page up and Page down keys now also work.
- The save as image option was enhanced to save Bitmaps (BMP) and GIF files. This is a great feature for those that want to publish their results on the web.
- The shareware trial period was increased to 30days (from 15).
- A single overall result figure is now calculated. We called this the PassMark Rating. It's a weighted average of all the other test results.
- A menu option was added to directly access the PassMark baseline database on the Web (you still need to be a registered user to use these additional baselines however).
V1.0a Released on the 3/June/99
Minor patch to correctly report speed for high end AMD CPUs.
V1.0 Released on the 16/May/99
First version released to the world. Wins numerous awards, whoopeee :-